Dataset Viewer
Auto-converted to Parquet Duplicate
id
stringlengths
13
13
domain
stringclasses
5 values
topic
stringclasses
5 values
risk_tier
stringclasses
7 values
user_prompt
stringlengths
219
368
thought_chain
stringlengths
254
643
chosen_response
stringlengths
470
1.36k
rejected_response
stringlengths
95
252
formal_proof
stringlengths
214
1k
dpo_critique
stringclasses
20 values
socratic_challenge
stringlengths
45
99
lemma_resolution
stringlengths
159
308
math_r1_00001
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
AIME / IMO HARD
Competition Benchmark: Lean 4 Mathlib Research [Problem #NT-PELL-01000] Domain: Olympiad Number Theory Problem Statement: Find all positive integer pairs $(x, y)$ such that $x^2 - 21y^2 = 1$ with $x \equiv 35 \pmod{43}$. Prove that there are infinitely many such pairs and determine the periodicity of solutions in $\mat...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{21}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 43 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(21\tilde{y}_1^2)$ is a quadratic non-...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{21}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 43 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(21\tilde{y}_1^2)$ is a quadratic non-...
Substitute $x=35$ directly into $x^2 - 21y^2 = 1$. Then $y = \sqrt{(35^2-1)/21}$. Since $35^2-1$ can be made divisible by $21$, solutions exist without checking group periodicity.
**Theorem & Complete Algebraic Number Theory Solution:** 1. **Pell Group Action**: Solutions satisfy $x_n + y_n\sqrt{21} = (x_1 + y_1\sqrt{21})^n$. 2. **Periodicity**: Over $\mathbb{F}_{43}$, if $(21/43) = 1$, period divides $p-1$; if $(21/43) = -1$, period divides $p+1$. 3. **Infinitude**: The residue $x_{n_0} \equiv...
Chosen proves periodicity in algebraic number field and analyzes quadratic character; Rejected naively evaluates a single value without proving infinitude or periodicity.
How does the norm map $N(\alpha) = 1$ in $\mathbb{F}_{41}$ force the period to divide $p+1$?
<thought>Norm map kernel has order $(p^2-1)/(p-1) = p+1$.</thought> **Lemma**: The cyclic kernel of $N: \mathbb{F}_{41}^\times \to \mathbb{F}_{43}^\times$ has exact cardinality $p+1$.
math_r1_00002
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
IMO SHORTLIST
Competition Benchmark: AIME 2026 Invitational [Problem #NT-LTE-01001] Domain: Olympiad Number Theory Problem Statement: Let $p = 17$ be an odd prime. Suppose positive integers $a, b$ satisfy $p \mid (a - b)$ but $p \nmid a$ and $p \nmid b$. Prove the Lifting The Exponent (LTE) identity for $n = 17$: \[ v_{17}(a^{17} - ...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
By expanding $a^n - b^n = (a-b)(a^{n-1} + \dots + b^{n-1})$, each of the $n$ terms is equal to $a^{n-1} \pmod p$, so the sum is $n a^{n-1}$. Thus $v_p(a^n - b^n) = v_p(a-b) + v_p(n a^{n-1}) = v_p(a-b) + v_p(n) + (n-1)v_p(a)$, which gives an extra term.
**Proof via Binomial p-adic Expansion:** 1. Let $v_p(a - b) = k \ge 1$. Write $a = b + c$ with $v_p(c) = k$. 2. Expanding $a^n = (b + c)^n = b^n + n b^{n-1} c + \sum_{j=2}^n \binom{n}{j} b^{n-j} c^j$. 3. For $j \ge 2$, $v_p(\binom{n}{j} c^j) \ge j k - v_p(j) > k + v_p(n)$ for odd prime $p$. 4. Therefore $v_p(a^n - b^n...
Chosen establishes p-adic valuations with strict inequality $j k - v_p(j) > k + v_p(n)$; Rejected incorrectly asserts $v_p(a) > 0$ when $p \nmid a$.
Why does LTE fail for $p = 2$ when $n$ is even?
<thought>For p=2 and j=2, 2k - 1 = k + (k-1) equals k when k=1, causing cancellation.</thought> **Resolution**: For $p=2$, the second-order binomial coefficient has $v_2(\binom{n}{2} c^2) = v_2(n) - 1 + 2k$, requiring the extra correction term $v_2(a+b)-1$.
math_r1_00003
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
PUTNAM / IMO HARD
Competition Benchmark: Lean 4 Mathlib Research [Problem #NT-GAUSS-01002] Domain: Olympiad Number Theory Problem Statement: Evaluate the Legendre symbol $\left(\frac{37}{71}\right)$ using Gauss's Law of Quadratic Reciprocity. Prove whether the congruence $x^2 \equiv 37 \pmod{71}$ is solvable in integers.
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 37 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $71 \pmod{37} = 34$. - Decompose into prime factors and apply Euler's criterion. </thou...
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 37 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $71 \pmod{37} = 34$. - Decompose into prime factors and apply Euler's criterion. </thou...
Since both 37 and 71 are odd primes, $(p/q) = -(q/p)$ always holds by Gauss's theorem. Evaluating directly gives the opposite sign.
**Evaluation via Quadratic Reciprocity:** 1. **Parity Check**: Since $p = 37 \equiv 1 \pmod 4$, $(-1)^{\frac{37-1}{2}\frac{71-1}{2}} = 1$. 2. **Reciprocity Application**: $\left(\frac{37}{71}\right) = \left(\frac{71}{37}\right)$. 3. **Modular Reduction**: Evaluating $\left(\frac{34}{37}\right) = 1$. Therefore, the co...
Chosen correctly analyzes $p \equiv 1 \pmod 4$ to obtain $(p/q) = +(q/p)$; Rejected misremembers Quadratic Reciprocity as always having a negative sign.
How does Gauss's Lemma with half-intervals prove quadratic reciprocity?
<thought>Counting lattice points in rectangle of size (p-1)/2 by (q-1)/2.</thought> **Resolution**: Eisenstein's geometric proof counts interior integer points in the rectangle $[1, (p-1)/2] \times [1, (q-1)/2]$.
math_r1_00004
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
AIME / USAMO
Competition Benchmark: Putnam Mathematical Competition [Problem #NT-CRT-01003] Domain: Olympiad Number Theory Problem Statement: Solve the simultaneous system of modular polynomial congruences: \[ x \equiv 14 \pmod{17}, \quad x \equiv 18 \pmod{19} \] and lift the unique solution modulo $M = 323$ to modulo $M^2 = 104329...
<thought> 1. CRT Inversion: - $M = 17 \times 19 = 323$. - $M_1 = 19, M_2 = 17$. - $y_1 = 17^{-1} \pmod{19} = 9$, $y_2 = 19^{-1} \pmod{17} = 9$. - $x_0 = (14 \cdot 19 \cdot 9 + 18 \cdot 17 \cdot 9) \pmod{323} = 303$. 2. Hensel Lifting Step: - For $f(x) = x - x_0 \equiv 0 \pmod M$, derivative $f'(x_0) = 1 ...
<thought> 1. CRT Inversion: - $M = 17 \times 19 = 323$. - $M_1 = 19, M_2 = 17$. - $y_1 = 17^{-1} \pmod{19} = 9$, $y_2 = 19^{-1} \pmod{17} = 9$. - $x_0 = (14 \cdot 19 \cdot 9 + 18 \cdot 17 \cdot 9) \pmod{323} = 303$. 2. Hensel Lifting Step: - For $f(x) = x - x_0 \equiv 0 \pmod M$, derivative $f'(x_0) = 1 ...
Multiply the two equations: $x^2 \equiv 252 \pmod{323}$. Then take square root $x = \sqrt{252}$, which is not integer-preserving.
**Exact CRT & Hensel Lifting:** 1. **BΓ©zout Multipliers**: $19 \cdot 9 \equiv 1 \pmod{17}$ and $17 \cdot 9 \equiv 1 \pmod{19}$. 2. **CRT Base Solution**: $x_0 \equiv 14(19)(9) + 18(17)(9) \equiv 303 \pmod{323}$. 3. **Hensel Lift**: Since $\gcd(f'(x_0), M) = 1$, the unique lift modulo $104329$ is \boxed{x \equiv 303 \p...
Chosen applies rigorous BΓ©zout modular inverses in CRT and verifies Hensel nonsingularity condition; Rejected commits an illegal non-linear multiplication of congruences.
What happens during Hensel lifting if $f'(x_0) \equiv 0 \pmod p$?
<thought>Branching into no solutions or p distinct solutions depending on f(x0) mod p^2.</thought> **Resolution**: If $f'(x_0) \equiv 0 \pmod p$, the lift exists if and only if $f(x_0) \equiv 0 \pmod{p^2}$, generating $p$ distinct lifts.
math_r1_00005
Higher Algebra & Functional Equations
Higher Algebra
PUTNAM / IMO HARD
Competition Benchmark: Putnam Mathematical Competition [Problem #ALG-CAUCHY-01004] Domain: Higher Algebra & Functional Equations Problem Statement: Find all functions $f: \mathbb{R} \to \mathbb{R}$ satisfying $f(x f(y) + f(x)) = 39 x f(y) + f(f(x))$ for all $x, y \in \mathbb{R}$.
<thought> 1. Trivial and Linear Ansatz: - $f(x) \equiv 0$ is valid. - For $f(x) = k x$, $k^2 xy + k^2 x = 39k xy + k^2 x \implies k = 39$. 2. Injectivity & Discontinuity Backtracking: - Suppose $f(y_1) = f(y_2)$. - $39 x f(y_1) + f(f(x)) = 39 x f(y_2) + f(f(x)) \implies f(y_1) = f(y_2)$ for $x \neq 0$. Thus...
<thought> 1. Trivial and Linear Ansatz: - $f(x) \equiv 0$ is valid. - For $f(x) = k x$, $k^2 xy + k^2 x = 39k xy + k^2 x \implies k = 39$. 2. Injectivity & Discontinuity Backtracking: - Suppose $f(y_1) = f(y_2)$. - $39 x f(y_1) + f(f(x)) = 39 x f(y_2) + f(f(x)) \implies f(y_1) = f(y_2)$ for $x \neq 0$. Thus...
Assume $f(x) = a x + b$. Then $a(ax(ay+b) + ax+b) + b = 39x(ay+b) + a(ax+b)+b$. Comparing coefficients gives $a=39, b=0$. No other solutions exist because all functions are polynomials.
**Analytic Classification:** 1. **Trivial Case**: $f(x) \equiv 0$. 2. **Injectivity**: If $f(y_1) = f(y_2)$, then for $x \neq 0$, $39 x f(y_1) + f(f(x)) = 39 x f(y_2) + f(f(x)) \implies f$ is injective. 3. **Hamel Elimination**: Multiplicativity $f(x^2) = \frac{1}{39} f(x)^2 \ge 0$ enforces positivity on $\mathbb{R}^+...
Chosen proves strict injectivity and eliminates non-measurable Hamel bases; Rejected assumes polynomial form without justification.
How does boundedness on a set of positive Lebesgue measure force linearity in Cauchy equations?
<thought>Steinhaus Theorem on difference sets A - A containing an interval around 0.</thought> **Resolution**: By Steinhaus Theorem, $A - A$ contains an open neighborhood of 0, ensuring continuity at 0 and linearity everywhere.
math_r1_00006
Higher Algebra & Functional Equations
Higher Algebra
AIME / USAMO
Competition Benchmark: AIME 2026 Invitational [Problem #ALG-VIETA-01005] Domain: Higher Algebra & Polynomials Problem Statement: Let $x_1, x_2, x_3$ be roots of $P(t) = t^3 - 2t^2 + 10t - 2 = 0$. Using Newton's sums $p_k = x_1^k + x_2^k + x_3^k$, compute $p_4$ exactly.
<thought> 1. Newton's Sums Recurrence: - $p_1 = e_1 = 2$. - $p_2 = e_1 p_1 - 2 e_2 = 2(2) - 2(10) = -16$. - $p_3 = e_1 p_2 - e_2 p_1 + 3 e_3 = 2(-16) - 10(2) + 3(2) = -46$. - $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 2(-46) - 10(-16) + 2(2) = 72$. </thought>
<thought> 1. Newton's Sums Recurrence: - $p_1 = e_1 = 2$. - $p_2 = e_1 p_1 - 2 e_2 = 2(2) - 2(10) = -16$. - $p_3 = e_1 p_2 - e_2 p_1 + 3 e_3 = 2(-16) - 10(2) + 3(2) = -46$. - $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 2(-46) - 10(-16) + 2(2) = 72$. </thought> **Calculation via Newton-Girard Identities:** 1. $p_...
Compute roots directly using Cardano's formula. Since $p_4 = (x_1+x_2+x_3)^4 = 2^4 = 16$, we obtain the answer without Newton identities.
**Calculation via Newton-Girard Identities:** 1. $p_1 = e_1 = 2$. 2. $p_2 = e_1 p_1 - 2e_2 = -16$. 3. $p_3 = e_1 p_2 - e_2 p_1 + 3e_3 = -46$. 4. $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 2(-46) - 10(-16) + 2(2) = \boxed{72}$$
Chosen applies Newton-Girard identities correctly subtracting cross-terms; Rejected equates $\sum x_i^4$ with $(\sum x_i)^4$.
How do Newton sums generalize to arbitrary degree $n$ polynomials?
<thought>Logarithmic derivative of polynomial generating function P'(t)/P(t).</thought> **Resolution**: Expanding $\frac{P'(t)}{P(t)} = \sum_{k=1}^\infty p_k t^{-k}$ generates the general Newton recurrence for all degrees.
math_r1_00007
Higher Algebra & Functional Equations
Higher Algebra
IMO SHORTLIST
Competition Benchmark: IMO Shortlist Benchmark [Problem #ALG-JENSEN-01006] Domain: Higher Algebra & Inequalities Problem Statement: For positive reals $a, b, c$ with $a + b + c = 1$, prove the strict convexity bound: \[ \frac{a}{\sqrt{1 - a}} + \frac{b}{\sqrt{1 - b}} + \frac{c}{\sqrt{1 - c}} \ge \frac{\sqrt{6}}{2} \]
<thought> 1. Convexity Verification: - Define $f(x) = \frac{x}{\sqrt{1-x}}$ for $x \in (0, 1)$. - $f'(x) = \frac{1 - x/2}{(1-x)^{3/2}}$. - $f''(x) = \frac{3(1-x/2)}{2(1-x)^{5/2}} - \frac{1/2}{(1-x)^{3/2}} = \frac{4 - x}{4(1-x)^{5/2}} > 0$. - Since $f''(x) > 0$, $f$ is strictly convex on $(0, 1)$. 2. Jensen ...
<thought> 1. Convexity Verification: - Define $f(x) = \frac{x}{\sqrt{1-x}}$ for $x \in (0, 1)$. - $f'(x) = \frac{1 - x/2}{(1-x)^{3/2}}$. - $f''(x) = \frac{3(1-x/2)}{2(1-x)^{5/2}} - \frac{1/2}{(1-x)^{3/2}} = \frac{4 - x}{4(1-x)^{5/2}} > 0$. - Since $f''(x) > 0$, $f$ is strictly convex on $(0, 1)$. 2. Jensen ...
Apply Cauchy-Schwarz directly: $(\sum a/\sqrt{1-a})(\sum a\sqrt{1-a}) \ge (\sum a)^2 = 1$. This gives lower bound 1, which is weaker than $\sqrt{6}/2$.
**Proof via Jensen's Inequality:** 1. **Convexity**: Let $f(x) = \frac{x}{\sqrt{1-x}}$. Second derivative $f''(x) = \frac{4 - x}{4(1-x)^{5/2}} > 0$ on $(0, 1)$, confirming strict convexity. 2. **Jensen's Bound**: \[ \frac{f(a) + f(b) + f(c)}{3} \ge f\left(\frac{a+b+c}{3}\right) = f\left(\frac{1}{3}\right) = \frac{1/3}...
Chosen establishes second-derivative convexity $f''(x) > 0$ and applies Jensen to hit the exact tight bound; Rejected uses a loose Cauchy-Schwarz.
How does Karamata's Majorization theorem generalize Jensen for sorted sequences?
<thought>Majorization x \succ y implies \sum f(x_i) \ge \sum f(y_i) for convex f.</thought> **Resolution**: Karamata proves that if $(x) \succ (y)$ in majorization order, $\sum f(x_i) \ge \sum f(y_i)$ for any convex function $f$.
math_r1_00008
Higher Algebra & Functional Equations
Higher Algebra
PUTNAM / RESEARCH
Competition Benchmark: Putnam Mathematical Competition [Problem #ALG-CHEBY-01007] Domain: Higher Algebra & Polynomials Problem Statement: Derive the closed-form expression for the Chebyshev polynomial of the first kind $T_{3}(x)$ satisfying $T_{3}(\cos \theta) = \cos(3\theta)$. Find all real roots of $T_{3}(x) = 0$ in ...
<thought> 1. Recurrence Relation: - $T_0(x) = 1, T_1(x) = x$. - $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$ from $\cos((n+1)\theta) + \cos((n-1)\theta) = 2\cos\theta\cos(n\theta)$. 2. Exact Roots: - $\cos(n\theta) = 0 \implies n\theta_k = \frac{2k-1}{2}\pi \implies \theta_k = \frac{2k-1}{2n}\pi$ for $k = 1, \dots, n...
<thought> 1. Recurrence Relation: - $T_0(x) = 1, T_1(x) = x$. - $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$ from $\cos((n+1)\theta) + \cos((n-1)\theta) = 2\cos\theta\cos(n\theta)$. 2. Exact Roots: - $\cos(n\theta) = 0 \implies n\theta_k = \frac{2k-1}{2}\pi \implies \theta_k = \frac{2k-1}{2n}\pi$ for $k = 1, \dots, n...
Since $T_n(x)$ is a polynomial of degree 3, its roots are equally spaced numbers between -1 and 1: $x_k = -1 + \frac{2k}{3}$, which ignores the non-linear cosine density.
**Derivation & Spectral Decomposition:** 1. **Recurrence**: $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$. 2. **Trigonometric Closed Form**: $T_{3}(x) = \cos(3 \arccos x)$ for $x \in [-1, 1]$. 3. **Roots**: Setting $\cos(3\theta) = 0$ gives $n$ distinct real roots: \[ \boxed{x_k = \cos\left(\frac{2k - 1}{2 \cdot 3} \pi\right)...
Chosen derives exact non-linear Chebyshev nodes $x_k = \cos(\frac{2k-1}{2n}\pi)$; Rejected assumes a naive uniform arithmetic progression.
Why do Chebyshev nodes minimize the Runge interpolation error phenomenon?
<thought>Equioscillation and minimization of the monic polynomial supremum norm ||P||_\infty = 2^{1-n}.</thought> **Resolution**: Chebyshev roots minimize the product $\prod |x - x_k|$, making the monic polynomial norm $\|T_n\|_\infty = 2^{1-n}$ minimal on $[-1, 1]$.
math_r1_00009
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-AMGM-01008] Domain: Formal Logic & Real Analysis Theorem Statement: Formalize and prove in Lean 4 that for non-negative reals $a, b$, $2\sqrt{ab} \le a + b$.
<thought> 1. Lean 4 Formalization Architecture: - Target Theorem: $\forall (a \, b : \mathbb{R}), 0 \le a \to 0 \le b \to 2 \cdot \sqrt{a \cdot b} \le a + b$. - Algebraic identity in Lean: $(a + b) - 2\sqrt{ab} = (\sqrt{a} - \sqrt{b})^2 \ge 0$. - Tactics: `have`, `calc`, `sq_nonneg`, `Real.sq_sqrt`, `linarith`...
<thought> 1. Lean 4 Formalization Architecture: - Target Theorem: $\forall (a \, b : \mathbb{R}), 0 \le a \to 0 \le b \to 2 \cdot \sqrt{a \cdot b} \le a + b$. - Algebraic identity in Lean: $(a + b) - 2\sqrt{ab} = (\sqrt{a} - \sqrt{b})^2 \ge 0$. - Tactics: `have`, `calc`, `sq_nonneg`, `Real.sq_sqrt`, `linarith`...
```lean4 theorem am_gm_two (a b : ℝ) (ha : 0 ≀ a) (hb : 0 ≀ b) : 2 * Real.sqrt (a * b) ≀ a + b := by have h := (a - b)^2 linarith ```
```lean4 import Mathlib.Data.Real.Basic import Mathlib.Data.Real.Sqrt import Mathlib.Tactic.Positivity import Mathlib.Tactic.Linarith /-- AM-GM Inequality for n = 2 in Lean 4 -/ theorem am_gm_two (a b : ℝ) (ha : 0 ≀ a) (hb : 0 ≀ b) : 2 * Real.sqrt (a * b) ≀ a + b := by have h_sq : 0 ≀ (Real.sqrt a - Real.sqrt b)...
Chosen provides a complete Mathlib4 calc chain with square root transformations; Rejected misuses $(a-b)^2$ which cannot eliminate square roots.
How do you prove strict equality condition $2\sqrt{ab} = a + b \iff a = b$?
<thought>By sq_eq_zero_iff and Real.sqrt_inj.</thought> **Resolution**: Equality holds when $(\sqrt{a} - \sqrt{b})^2 = 0 \iff \sqrt{a} = \sqrt{b} \iff a = b$.
math_r1_00010
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-BERNOULLI-01009] Domain: Formal Logic & Analysis Theorem Statement: Formalize and prove Bernoulli's Inequality in Lean 4 for all $n : \mathbb{N}$ and $x \ge -1$: \[ (1 + x)^n \ge 1 + n x \]
<thought> 1. Mathematical Induction in Lean 4: - Base case n = 0: (1+x)^0 = 1 \ge 1 + 0 = 1. Trivial by ring / norm_num. - Inductive step: Assume (1+x)^k \ge 1 + kx. - Multiply by (1+x) \ge 0: (1+x)^{k+1} \ge (1 + kx)(1 + x) = 1 + (k+1)x + k x^2 \ge 1 + (k+1)x. - Tactics: induction', linarith, nlinarith, mu...
<thought> 1. Mathematical Induction in Lean 4: - Base case n = 0: (1+x)^0 = 1 \ge 1 + 0 = 1. Trivial by ring / norm_num. - Inductive step: Assume (1+x)^k \ge 1 + kx. - Multiply by (1+x) \ge 0: (1+x)^{k+1} \ge (1 + kx)(1 + x) = 1 + (k+1)x + k x^2 \ge 1 + (k+1)x. - Tactics: induction', linarith, nlinarith, mu...
```lean4 theorem bernoulli_inequality (x : ℝ) (n : β„•) : 1 + (n : ℝ) * x ≀ (1 + x) ^ n := by induction' n with k hk Β· simp Β· nlinarith ```
```lean4 import Mathlib.Data.Real.Basic import Mathlib.Tactic.Linarith import Mathlib.Tactic.Ring /-- Bernoulli's Inequality by induction in Lean 4 -/ theorem bernoulli_inequality (x : ℝ) (hx : -1 ≀ x) (n : β„•) : 1 + (n : ℝ) * x ≀ (1 + x) ^ n := by induction' n with k hk Β· simp Β· have h_pos : 0 ≀ 1 + x := by ...
Chosen establishes $0 \le 1+x$ positivity and proves inductive step via calc; Rejected omits hypothesis $x \ge -1$ making the theorem false for negative odd powers.
Why is the hypothesis $x \ge -1$ essential in the Lean 4 proof?
<thought>Multiplying inequalities by a negative term reverses the inequality sign.</thought> **Resolution**: If $1+x < 0$, `mul_le_mul_of_nonneg_right` fails because multiplication by a negative value reverses $\le$ to $\ge$.
math_r1_00011
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-SQRT2-01010] Domain: Formal Logic & Number Theory Theorem Statement: Formalize and prove in Lean 4 that $\sqrt{2}$ is irrational by showing there exist no coprime natural numbers $p, q$ with $p^2 = 2q^2$.
<thought> 1. Classical Parity Proof in Lean 4: - Target: `theorem sqrt_two_irrational (p q : β„•) (hq : q β‰  0) (h_coprime : Nat.Coprime p q) : p ^ 2 β‰  2 * q ^ 2`. - If $p^2 = 2q^2$, then $2 \mid p^2 \implies 2 \mid p$. - Write $p = 2k$, then $4k^2 = 2q^2 \implies q^2 = 2k^2 \implies 2 \mid q$. - Contradiction...
<thought> 1. Classical Parity Proof in Lean 4: - Target: `theorem sqrt_two_irrational (p q : β„•) (hq : q β‰  0) (h_coprime : Nat.Coprime p q) : p ^ 2 β‰  2 * q ^ 2`. - If $p^2 = 2q^2$, then $2 \mid p^2 \implies 2 \mid p$. - Write $p = 2k$, then $4k^2 = 2q^2 \implies q^2 = 2k^2 \implies 2 \mid q$. - Contradiction...
```lean4 theorem sqrt_two_irrational (p q : β„•) : p ^ 2 β‰  2 * q ^ 2 := by intro h have h2 : p = 0 := by linarith sorry ```
```lean4 import Mathlib.Data.Nat.Prime import Mathlib.Tactic.Ring /-- Proof of irrationality of sqrt(2) in Lean 4 -/ theorem sqrt_two_irrational (p q : β„•) (h_coprime : p.Coprime q) : p ^ 2 β‰  2 * q ^ 2 := by intro h_eq have h_even_p2 : 2 ∣ p ^ 2 := by use q ^ 2 exact h_eq.symm have h_even_p : 2 ∣ p :=...
Chosen uses formal Mathlib4 `Nat.Prime.dvd_of_dvd_pow` to prove parity contradiction on coprime gcd; Rejected collapses to sorry.
How does this generalize to $\sqrt{n}$ for any non-square integer $n$?
<thought>By p-adic valuation v_p(p^2) = 2 v_p(p) is always even, while v_p(n q^2) is odd for prime p with odd exponent in n.</thought> **Resolution**: Fundamental Theorem of Arithmetic implies $v_p(p^2)$ is even, whereas $v_p(n q^2) = v_p(n) + 2v_p(q)$ is odd, yielding an immediate contradiction.
math_r1_00012
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-LAGRANGE-01011] Domain: Formal Logic & Abstract Algebra Theorem Statement: Formalize in Lean 4 the coset partition equality: for any finite group $G$ and subgroup $H \le G$, $|G| = [G : H] \cdot |H|$.
<thought> 1. Lean 4 Abstract Algebra Architecture: - Typeclasses: `Group G`, `Fintype G`, `Subgroup H`. - Left coset equivalence relation $a \sim b \iff a^{-1}b \in H$. - Bijections between distinct cosets: $x H \cong H$ via $h \mapsto x h$. </thought>
<thought> 1. Lean 4 Abstract Algebra Architecture: - Typeclasses: `Group G`, `Fintype G`, `Subgroup H`. - Left coset equivalence relation $a \sim b \iff a^{-1}b \in H$. - Bijections between distinct cosets: $x H \cong H$ via $h \mapsto x h$. </thought> ```lean4 import Mathlib.Algebra.Group.Subgroup.Basic impo...
```lean4 theorem lagrange_group_order {G : Type*} [Group G] (H : Subgroup G) : Fintype.card G = Fintype.card H := by rfl ```
```lean4 import Mathlib.Algebra.Group.Subgroup.Basic import Mathlib.GroupTheory.Index import Mathlib.Data.Fintype.Card /-- Lagrange's Theorem for Finite Groups in Lean 4 -/ theorem lagrange_group_order {G : Type*} [Group G] [Fintype G] (H : Subgroup G) [Fintype H] : Fintype.card G = H.index * Fintype.card H := by ...
Chosen applies correct Mathlib4 subgroup quotient cardinality theorem; Rejected falsely claims subgroup order equals group order by rfl.
What is the corollary for prime-order groups?
<thought>If |G| = p, proper subgroups have order 1 or p, forcing G to be cyclic.</thought> **Resolution**: Every group of prime order $p$ has no non-trivial proper subgroups, and is therefore cyclic ($G \cong \mathbb{Z}/p\mathbb{Z}$).
math_r1_00013
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
PUTNAM / RESEARCH
Competition Benchmark: Putnam Mathematical Competition [Problem #COMB-RAMSEY-01012] Domain: Advanced Combinatorics Problem Statement: In complete graph $K_n$, prove using the probabilistic method with ErdΕ‘s expectation and the Symmetric LovΓ‘sz Local Lemma (LLL) that diagonal Ramsey number satisfies $R(3, 3) > \frac{\sq...
<thought> 1. Random Edge Coloring: - Each edge colored Red/Blue with $p = 1/2$. - Probability of monochromatic $K_k$: $P(A_S) = 2^{1 - \binom{k}{2}}$. 2. Dependency Graph & LLL: - Two subsets $S, T$ are dependent iff $|S \cap T| \ge 2$. - Degree $d = \binom{k}{2} \binom{n-2}{k-2}$. - LLL Condition: $e \c...
<thought> 1. Random Edge Coloring: - Each edge colored Red/Blue with $p = 1/2$. - Probability of monochromatic $K_k$: $P(A_S) = 2^{1 - \binom{k}{2}}$. 2. Dependency Graph & LLL: - Two subsets $S, T$ are dependent iff $|S \cap T| \ge 2$. - Degree $d = \binom{k}{2} \binom{n-2}{k-2}$. - LLL Condition: $e \c...
Color alternating edges Red and Blue. In any graph, there are at most $n/2$ edges of one color, so no monochromatic clique of size 3 can exist for $n < 2^3$.
**Probabilistic Bound via Symmetric LovΓ‘sz Local Lemma:** 1. **Bad Event Probability**: For subset $S$ of size $3$, $P(A_S) = 2^{1 - \binom{3}{2}}$. 2. **Dependency Degree**: $d = \binom{3}{2} \binom{n-2}{3-2} < \frac{3^2}{2} \frac{n^{3-2}}{(3-2)!}$. 3. **LLL Criterion**: $e \cdot P(A_S) \cdot (d + 1) \le 1$ ensures $...
Chosen derives the exact dependency degree $d$ and invokes Symmetric LLL; Rejected makes an invalid deterministic alternating coloring claim.
How does Spencer's polynomial method compare with LLL?
<thought>Spencer's fractional coloring achieves optimal constants within the probabilistic method.</thought> **Resolution**: Spencer's theorem matches the asymptotic constant while providing an algorithmic Moser-Tardos resample algorithm.
math_r1_00014
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
AIME / USAMO
Competition Benchmark: USAMO National Olympiad [Problem #COMB-DERANGE-01013] Domain: Advanced Combinatorics Problem Statement: Using the Principle of Inclusion-Exclusion, derive the exact closed formula for derangements $D_n = !9$ (permutations of $9$ elements with no fixed points) and evaluate $D_{9}$ exactly.
<thought> 1. Inclusion-Exclusion Formulation: - Let $A_i$ be permutations fixing $i$. $|A_i| = (n-1)!$. - Intersection of $k$ sets: $|igcap_{j=1}^k A_{i_j}| = (n-k)!$. - $D_n = n! - \sum |A_i| + \sum |A_i \cap A_j| - \dots = n! \sum_{k=0}^n \frac{(-1)^k}{k!}$. 2. Evaluation for n = 9: - $D_{n} = \lfloor \f...
<thought> 1. Inclusion-Exclusion Formulation: - Let $A_i$ be permutations fixing $i$. $|A_i| = (n-1)!$. - Intersection of $k$ sets: $|igcap_{j=1}^k A_{i_j}| = (n-k)!$. - $D_n = n! - \sum |A_i| + \sum |A_i \cap A_j| - \dots = n! \sum_{k=0}^n \frac{(-1)^k}{k!}$. 2. Evaluation for n = 9: - $D_{n} = \lfloor \f...
Since each element has $n-1$ choices for its position, $D_n = (n-1)^n$. For $n=9$, this gives $(9-1)^{9}$, which ignores dependency among cycles.
**Derivation via Inclusion-Exclusion:** 1. **Union of Fixed Points**: $|\bigcup_{i=1}^{9} A_i| = \sum_{k=1}^{9} (-1)^{k-1} \binom{9}{k} (9-k)! = 9! \sum_{k=1}^{9} \frac{(-1)^{k-1}}{k!}$. 2. **Complement**: $D_{9} = 9! \left(1 - \sum_{k=1}^{9} \frac{(-1)^{k-1}}{k!}\right) = 9! \sum_{k=0}^{9} \frac{(-1)^k}{k!}$. \boxed...
Chosen uses formal PIE with alternating sum converging to $n!/e$; Rejected uses independent choices $(n-1)^n$ allowing duplicate assignments.
How does the recurrence $D_n = (n-1)(D_{n-1} + D_{n-2})$ relate to permutation cycle decomposition?
<thought>Element 1 swaps with element k (D_{n-2}) or is part of a larger cycle (D_{n-1}).</thought> **Resolution**: Partitioning based on whether element 1 belongs to a 2-cycle or cycle of length $\ge 3$ yields $D_n = (n-1)(D_{n-1} + D_{n-2})$.
math_r1_00015
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
PUTNAM / IMO HARD
Competition Benchmark: IMO Shortlist Benchmark [Problem #COMB-CAYLEY-01014] Domain: Advanced Combinatorics Problem Statement: Prove Cayley's Formula for the number of labeled trees on $10$ vertices $T_{10} = 10^{10-2}$ by establishing a bijective correspondence with PrΓΌfer sequences of length $10-2$.
<thought> 1. PrΓΌfer Sequence Bijection: - Tree \to PrΓΌfer: Successively remove leaf with smallest label and record its neighbor. Repeat until 2 vertices remain. - Result is a sequence of length n - 2 with entries in \{1, \dots, n\}. - Number of such sequences is n^{n-2}. - Inverse algorithm reconstructs uni...
<thought> 1. PrΓΌfer Sequence Bijection: - Tree \to PrΓΌfer: Successively remove leaf with smallest label and record its neighbor. Repeat until 2 vertices remain. - Result is a sequence of length n - 2 with entries in \{1, \dots, n\}. - Number of such sequences is n^{n-2}. - Inverse algorithm reconstructs uni...
A tree has $n-1$ edges chosen from $\binom{n}{2}$ possible edges. So $T_n = \binom{\binom{n}{2}}{n-1}$, which counts all acyclic and cyclic subgraphs indiscriminately.
**Bijective Proof via PrΓΌfer Sequences:** 1. **Tree to Sequence**: Iteratively delete the minimal leaf $v$ and append its unique neighbor $u$ to sequence $P$. The process terminates at length $10-2$. 2. **Reconstruction**: Each vertex appears in $P$ exactly $\text{deg}(v) - 1$ times. The minimal missing label is the f...
Chosen establishes explicit bijection with PrΓΌfer sequences; Rejected naively counts arbitrary subsets of edges $\binom{N}{n-1}$.
How does Matrix-Tree Theorem prove Cayley's formula algebraically?
<thought>Laplacian matrix of K_n has eigenvalues n with multiplicity n-1, so det cofactor is n^{n-2}.</thought> **Resolution**: The Laplacian of $K_n$ has spectrum $\{0, n, \dots, n\}$, and any cofactor equals $\frac{1}{n} \prod_{i=1}^{n-1} n = n^{n-2}$.
math_r1_00016
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
AIME / USAMO
Competition Benchmark: IMO Shortlist Benchmark [Problem #COMB-CATALAN-01015] Domain: Advanced Combinatorics Problem Statement: Using Ordinary Generating Functions (OGF), solve the Catalan recurrence $C_0 = 1, C_{n+1} = \sum_{i=0}^n C_i C_{n-i}$ and compute $C_{5}$.
<thought> 1. OGF Setup: - Let C(x) = \sum_{n=0}^\infty C_n x^n. - C(x) = 1 + x C(x)^2 \implies x C(x)^2 - C(x) + 1 = 0. - Roots: C(x) = \frac{1 \pm \sqrt{1 - 4x}}{2x}. - Since C(0) = 1, we select minus sign: C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}. 2. Taylor Expansion: - \sqrt{1-4x} = 1 - \sum_{n=1}^\infty \...
<thought> 1. OGF Setup: - Let C(x) = \sum_{n=0}^\infty C_n x^n. - C(x) = 1 + x C(x)^2 \implies x C(x)^2 - C(x) + 1 = 0. - Roots: C(x) = \frac{1 \pm \sqrt{1 - 4x}}{2x}. - Since C(0) = 1, we select minus sign: C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}. 2. Taylor Expansion: - \sqrt{1-4x} = 1 - \sum_{n=1}^\infty \...
The recurrence $C_{n+1} = \sum C_i C_{n-i}$ is solved by $C_n = 2^n$ because at each step we make a binary choice.
**Closed Form via Algebraic Generating Function:** 1. **Functional Equation**: $C(x) = 1 + x C(x)^2 \implies C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}$. 2. **Binomial Series**: $\sqrt{1 - 4x} = 1 - 2 \sum_{n=1}^\infty \frac{1}{n} \binom{2n-2}{n-1} x^n$. 3. **Extraction**: $C_n = \frac{1}{n+1} \binom{2n}{n}$. \boxed{C_{5} =...
Chosen derives algebraic OGF $C(x) = (1-\sqrt{1-4x})/2x$ and extracts exact coefficients; Rejected mistakes Catalan growth for exponential $2^n$.
How does Dyck path reflection principle prove $C_n = \binom{2n}{n} - \binom{2n}{n+1}$?
<thought>Reflecting bad paths that touch y = -1 across the line y = -1 establishes a bijection to paths ending at (2n, -2).</thought> **Resolution**: AndrΓ©'s Reflection Principle reflects the portion of an invalid path after its first touch of $y=-1$, mapping invalid paths bijectively to $\binom{2n}{n-1}$.
math_r1_00017
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: AIME 2026 Invitational [Problem #GEOM-EULER-01016] Domain: Euclidean Geometry Problem Statement: In triangle $\triangle ABC$, inradius is $r = 27$ and circumradius is $R = 87$. Prove Euler's formula $d^2 = R(R - 2r)$ for distance $d = IO$ and compute $d$.
<thought> 1. Geometric Invariant: - Power of incenter $I$ with respect to circumcircle: $R^2 - d^2 = AI \cdot IM$. - Trillium Theorem: $IM = MB$. - $AI = r / \sin(A/2)$, $MB = 2R \sin(A/2)$. - $AI \cdot MB = 2Rr \implies d^2 = R(R - 2r) = 2871$. </thought>
<thought> 1. Geometric Invariant: - Power of incenter $I$ with respect to circumcircle: $R^2 - d^2 = AI \cdot IM$. - Trillium Theorem: $IM = MB$. - $AI = r / \sin(A/2)$, $MB = 2R \sin(A/2)$. - $AI \cdot MB = 2Rr \implies d^2 = R(R - 2r) = 2871$. </thought> **Synthetic Proof via Trillium Lemma:** 1. **Powe...
By Pythagoras theorem in $\triangle AIO$, $d^2 = R^2 + r^2$. For $R=87, r=27$, $d = \sqrt{8298}$, which ignores non-orthogonality.
**Synthetic Proof via Trillium Lemma:** 1. **Power of Incenter**: $R^2 - d^2 = AI \cdot IM$. 2. **Trillium Lemma**: $IM = MB = 2R \sin(A/2)$ and $AI = \frac{r}{\sin(A/2)}$. 3. **Product**: $AI \cdot IM = 2Rr \implies \boxed{d^2 = R(R - 2r) = 2871 \implies d = \sqrt{2871}}$.
Chosen applies Trillium Lemma and Power of Point on circumcircle; Rejected falsely assumes $OI \perp AI$ applying Pythagorean theorem.
How does $d^2 \ge 0$ immediately prove Euler's inequality $R \ge 2r$?
<thought>Real Euclidean distance requires d^2 >= 0, so R(R-2r) >= 0.</thought> **Resolution**: Since $R > 0$, $R(R-2r) = d^2 \ge 0$ yields $R \ge 2r$, with equality $R = 2r$ if and only if $d = 0$ (equilateral triangle).
math_r1_00018
Euclidean & Projective Geometry
Euclidean
USAMO / IMO
Competition Benchmark: Putnam Mathematical Competition [Problem #GEOM-PTOLEMY-01017] Domain: Euclidean & Projective Geometry Problem Statement: In cyclic quadrilateral $ABCD$ with sides $AB=8, BC=4, CD=7, DA=7$, prove Ptolemy's Theorem $AC \cdot BD = AB \cdot CD + BC \cdot DA$ using circular inversion.
<thought> 1. Geometric Inversion Method: - Invert with center A and arbitrary radius R. - Circle passing through A, B, C, D maps to a straight line containing B', C', D' in order. - Segment lengths under inversion: X'Y' = \frac{R^2 \cdot XY}{AX \cdot AY}. - Collinear segments: B'D' = B'C' + C'D'. - Subst...
<thought> 1. Geometric Inversion Method: - Invert with center A and arbitrary radius R. - Circle passing through A, B, C, D maps to a straight line containing B', C', D' in order. - Segment lengths under inversion: X'Y' = \frac{R^2 \cdot XY}{AX \cdot AY}. - Collinear segments: B'D' = B'C' + C'D'. - Subst...
Apply law of cosines to $\triangle ABC$ and $\triangle ADC$. Then add the two diagonal formulas directly without matching opposite supplementary angles.
**Proof via Circular Inversion:** 1. **Inversion Action**: Invert the figure at center $A$. The circumcircle through $A$ maps to a straight line on which points $B', C', D'$ lie in collinear order. 2. **Metric Transformation**: $B'D' = \frac{R^2 \cdot BD}{AB \cdot AD}$, $B'C' = \frac{R^2 \cdot BC}{AB \cdot AC}$, $C'D'...
Chosen establishes rigorous proof via circular inversion mapping circles to lines; Rejected fails to eliminate cosines of opposite angles.
How does Ptolemy's inequality hold for non-cyclic quadrilaterals?
<thought>Triangle inequality on the inverted points B', C', D' gives B'D' <= B'C' + C'D'.</thought> **Resolution**: For non-cyclic quadrilaterals, $B', C', D'$ form a non-degenerate triangle, so triangle inequality yields strict $AC \cdot BD < AB \cdot CD + BC \cdot DA$.
math_r1_00019
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: USAMO National Olympiad [Problem #GEOM-RADICAL-01018] Domain: Euclidean Geometry Problem Statement: Let $\omega_1, \omega_2, \omega_3$ be three non-concentric circles whose centers are not collinear. Prove that the three radical axes pairwise intersect at a unique point $P$ (the Radical Center).
<thought> 1. Power of a Point & Radical Axis Definition: - For circle \omega(O, R), \text{Pow}(X, \omega) = d(X, O)^2 - R^2. - Radical axis L_{12} = \{X : \text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)\}. 2. Intersection Proof: - Let P = L_{12} \cap L_{23}. - Then \text{Pow}(P, \omega_1) = \text{Pow}(P, ...
<thought> 1. Power of a Point & Radical Axis Definition: - For circle \omega(O, R), \text{Pow}(X, \omega) = d(X, O)^2 - R^2. - Radical axis L_{12} = \{X : \text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)\}. 2. Intersection Proof: - Let P = L_{12} \cap L_{23}. - Then \text{Pow}(P, \omega_1) = \text{Pow}(P, ...
Since all circles are round, their chords intersect at the center of mass of the three centers.
**Proof of Radical Center Concurrence:** 1. **Linearity of Radical Axis**: The locus $\text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)$ is a straight line perpendicular to the line of centers $O_1 O_2$. 2. **Intersection**: Since centers are non-collinear, line $L_{12}$ and $L_{23}$ intersect at unique point $P$. 3. ...
Chosen applies analytic Power of Point transitivity $\text{Pow}(P, \omega_1) = \text{Pow}(P, \omega_3)$; Rejected makes a nonsensical center-of-mass heuristic.
What happens if the centers $O_1, O_2, O_3$ are collinear?
<thought>The radical axes are parallel lines perpendicular to the line of centers.</thought> **Resolution**: If centers are collinear, the three radical axes are mutually parallel perpendiculars (intersecting at infinity in projective plane).
math_r1_00020
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: Lean 4 Mathlib Research [Problem #GEOM-SIMSON-01019] Domain: Euclidean & Projective Geometry Problem Statement: Prove the Simson Line Theorem: The orthogonal projections $P_a, P_b, P_c$ of point $P$ onto sides $BC, CA, AB$ of $\triangle ABC$ are collinear if and only if $P$ lies on the circumcirc...
<thought> 1. Cyclic Quadrilateral Angle Chasing: - Points P, P_a, B, P_c are concyclic on circle with diameter PB. - Points P, P_a, C, P_b are concyclic on circle with diameter PC. - Collinearity of P_a, P_b, P_c requires \angle P P_a P_c = \angle P P_a P_b. - In cyclic quads: \angle P P_a P_c = \angle P B ...
<thought> 1. Cyclic Quadrilateral Angle Chasing: - Points P, P_a, B, P_c are concyclic on circle with diameter PB. - Points P, P_a, C, P_b are concyclic on circle with diameter PC. - Collinearity of P_a, P_b, P_c requires \angle P P_a P_c = \angle P P_a P_b. - In cyclic quads: \angle P P_a P_c = \angle P B ...
Connect the three pedal points to form $\triangle P_a P_b P_c$. Use Heron's formula for area. Setting area to 0 gives $P_a = P_b$, which requires $P$ to be at a vertex.
**Angle-Chasing Proof of Simson's Theorem:** 1. **Concyclic Pedal Quadrilaterals**: Since $\angle P P_a B = \angle P P_c B = 90^\circ$, quadrilateral $P P_a B P_c$ is cyclic with diameter $PB$. Likewise, $P P_a C P_b$ is cyclic with diameter $PC$. 2. **Inscribed Angle Equivalence**: $\angle P P_a P_c = \angle P B A$ a...
Chosen uses cyclic pedal circles to prove $\angle PBA = \angle PCA$ iff $P$ on circumcircle; Rejected confuses collinearity with point collapse $P_a = P_b$.
How does the Simson line bisect the segment $P H$ where $H$ is the orthocenter?
<thought>Homothety of ratio 1/2 from H maps circumcircle to nine-point circle, placing midpoint of PH on Simson line.</thought> **Resolution**: The midpoint of $PH$ lies on the nine-point circle and the Simson line of $P$, meaning the Simson line bisects $PH$.
math_r1_00021
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
AIME / IMO HARD
Competition Benchmark: IMO Shortlist Benchmark [Problem #NT-PELL-01020] Domain: Olympiad Number Theory Problem Statement: Find all positive integer pairs $(x, y)$ such that $x^2 - 2y^2 = 1$ with $x \equiv 93 \pmod{101}$. Prove that there are infinitely many such pairs and determine the periodicity of solutions in $\mat...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{2}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 101 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(2\tilde{y}_1^2)$ is a quadratic non-r...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{2}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 101 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(2\tilde{y}_1^2)$ is a quadratic non-r...
Substitute $x=93$ directly into $x^2 - 2y^2 = 1$. Then $y = \sqrt{(93^2-1)/2}$. Since $93^2-1$ can be made divisible by $2$, solutions exist without checking group periodicity.
**Theorem & Complete Algebraic Number Theory Solution:** 1. **Pell Group Action**: Solutions satisfy $x_n + y_n\sqrt{2} = (x_1 + y_1\sqrt{2})^n$. 2. **Periodicity**: Over $\mathbb{F}_{101}$, if $(2/101) = 1$, period divides $p-1$; if $(2/101) = -1$, period divides $p+1$. 3. **Infinitude**: The residue $x_{n_0} \equiv ...
Chosen proves periodicity in algebraic number field and analyzes quadratic character; Rejected naively evaluates a single value without proving infinitude or periodicity.
How does the norm map $N(\alpha) = 1$ in $\mathbb{F}_{103}$ force the period to divide $p+1$?
<thought>Norm map kernel has order $(p^2-1)/(p-1) = p+1$.</thought> **Lemma**: The cyclic kernel of $N: \mathbb{F}_{103}^\times \to \mathbb{F}_{101}^\times$ has exact cardinality $p+1$.
math_r1_00022
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
IMO SHORTLIST
Competition Benchmark: Putnam Mathematical Competition [Problem #NT-LTE-01021] Domain: Olympiad Number Theory Problem Statement: Let $p = 7$ be an odd prime. Suppose positive integers $a, b$ satisfy $p \mid (a - b)$ but $p \nmid a$ and $p \nmid b$. Prove the Lifting The Exponent (LTE) identity for $n = 11$: \[ v_{7}(a^...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
By expanding $a^n - b^n = (a-b)(a^{n-1} + \dots + b^{n-1})$, each of the $n$ terms is equal to $a^{n-1} \pmod p$, so the sum is $n a^{n-1}$. Thus $v_p(a^n - b^n) = v_p(a-b) + v_p(n a^{n-1}) = v_p(a-b) + v_p(n) + (n-1)v_p(a)$, which gives an extra term.
**Proof via Binomial p-adic Expansion:** 1. Let $v_p(a - b) = k \ge 1$. Write $a = b + c$ with $v_p(c) = k$. 2. Expanding $a^n = (b + c)^n = b^n + n b^{n-1} c + \sum_{j=2}^n \binom{n}{j} b^{n-j} c^j$. 3. For $j \ge 2$, $v_p(\binom{n}{j} c^j) \ge j k - v_p(j) > k + v_p(n)$ for odd prime $p$. 4. Therefore $v_p(a^n - b^n...
Chosen establishes p-adic valuations with strict inequality $j k - v_p(j) > k + v_p(n)$; Rejected incorrectly asserts $v_p(a) > 0$ when $p \nmid a$.
Why does LTE fail for $p = 2$ when $n$ is even?
<thought>For p=2 and j=2, 2k - 1 = k + (k-1) equals k when k=1, causing cancellation.</thought> **Resolution**: For $p=2$, the second-order binomial coefficient has $v_2(\binom{n}{2} c^2) = v_2(n) - 1 + 2k$, requiring the extra correction term $v_2(a+b)-1$.
math_r1_00023
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
PUTNAM / IMO HARD
Competition Benchmark: Lean 4 Mathlib Research [Problem #NT-GAUSS-01022] Domain: Olympiad Number Theory Problem Statement: Evaluate the Legendre symbol $\left(\frac{17}{47}\right)$ using Gauss's Law of Quadratic Reciprocity. Prove whether the congruence $x^2 \equiv 17 \pmod{47}$ is solvable in integers.
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 17 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $47 \pmod{17} = 13$. - Decompose into prime factors and apply Euler's criterion. </thou...
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 17 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $47 \pmod{17} = 13$. - Decompose into prime factors and apply Euler's criterion. </thou...
Since both 17 and 47 are odd primes, $(p/q) = -(q/p)$ always holds by Gauss's theorem. Evaluating directly gives the opposite sign.
**Evaluation via Quadratic Reciprocity:** 1. **Parity Check**: Since $p = 17 \equiv 1 \pmod 4$, $(-1)^{\frac{17-1}{2}\frac{47-1}{2}} = 1$. 2. **Reciprocity Application**: $\left(\frac{17}{47}\right) = \left(\frac{47}{17}\right)$. 3. **Modular Reduction**: Evaluating $\left(\frac{13}{17}\right) = 1$. Therefore, the co...
Chosen correctly analyzes $p \equiv 1 \pmod 4$ to obtain $(p/q) = +(q/p)$; Rejected misremembers Quadratic Reciprocity as always having a negative sign.
How does Gauss's Lemma with half-intervals prove quadratic reciprocity?
<thought>Counting lattice points in rectangle of size (p-1)/2 by (q-1)/2.</thought> **Resolution**: Eisenstein's geometric proof counts interior integer points in the rectangle $[1, (p-1)/2] \times [1, (q-1)/2]$.
math_r1_00024
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
AIME / USAMO
Competition Benchmark: Putnam Mathematical Competition [Problem #NT-CRT-01023] Domain: Olympiad Number Theory Problem Statement: Solve the simultaneous system of modular polynomial congruences: \[ x \equiv 16 \pmod{17}, \quad x \equiv 4 \pmod{19} \] and lift the unique solution modulo $M = 323$ to modulo $M^2 = 104329$...
<thought> 1. CRT Inversion: - $M = 17 \times 19 = 323$. - $M_1 = 19, M_2 = 17$. - $y_1 = 17^{-1} \pmod{19} = 9$, $y_2 = 19^{-1} \pmod{17} = 9$. - $x_0 = (16 \cdot 19 \cdot 9 + 4 \cdot 17 \cdot 9) \pmod{323} = 118$. 2. Hensel Lifting Step: - For $f(x) = x - x_0 \equiv 0 \pmod M$, derivative $f'(x_0) = 1 \...
<thought> 1. CRT Inversion: - $M = 17 \times 19 = 323$. - $M_1 = 19, M_2 = 17$. - $y_1 = 17^{-1} \pmod{19} = 9$, $y_2 = 19^{-1} \pmod{17} = 9$. - $x_0 = (16 \cdot 19 \cdot 9 + 4 \cdot 17 \cdot 9) \pmod{323} = 118$. 2. Hensel Lifting Step: - For $f(x) = x - x_0 \equiv 0 \pmod M$, derivative $f'(x_0) = 1 \...
Multiply the two equations: $x^2 \equiv 64 \pmod{323}$. Then take square root $x = \sqrt{64}$, which is not integer-preserving.
**Exact CRT & Hensel Lifting:** 1. **BΓ©zout Multipliers**: $19 \cdot 9 \equiv 1 \pmod{17}$ and $17 \cdot 9 \equiv 1 \pmod{19}$. 2. **CRT Base Solution**: $x_0 \equiv 16(19)(9) + 4(17)(9) \equiv 118 \pmod{323}$. 3. **Hensel Lift**: Since $\gcd(f'(x_0), M) = 1$, the unique lift modulo $104329$ is \boxed{x \equiv 118 \pm...
Chosen applies rigorous BΓ©zout modular inverses in CRT and verifies Hensel nonsingularity condition; Rejected commits an illegal non-linear multiplication of congruences.
What happens during Hensel lifting if $f'(x_0) \equiv 0 \pmod p$?
<thought>Branching into no solutions or p distinct solutions depending on f(x0) mod p^2.</thought> **Resolution**: If $f'(x_0) \equiv 0 \pmod p$, the lift exists if and only if $f(x_0) \equiv 0 \pmod{p^2}$, generating $p$ distinct lifts.
math_r1_00025
Higher Algebra & Functional Equations
Higher Algebra
PUTNAM / IMO HARD
Competition Benchmark: Lean 4 Mathlib Research [Problem #ALG-CAUCHY-01024] Domain: Higher Algebra & Functional Equations Problem Statement: Find all functions $f: \mathbb{R} \to \mathbb{R}$ satisfying $f(x f(y) + f(x)) = 34 x f(y) + f(f(x))$ for all $x, y \in \mathbb{R}$.
<thought> 1. Trivial and Linear Ansatz: - $f(x) \equiv 0$ is valid. - For $f(x) = k x$, $k^2 xy + k^2 x = 34k xy + k^2 x \implies k = 34$. 2. Injectivity & Discontinuity Backtracking: - Suppose $f(y_1) = f(y_2)$. - $34 x f(y_1) + f(f(x)) = 34 x f(y_2) + f(f(x)) \implies f(y_1) = f(y_2)$ for $x \neq 0$. Thus...
<thought> 1. Trivial and Linear Ansatz: - $f(x) \equiv 0$ is valid. - For $f(x) = k x$, $k^2 xy + k^2 x = 34k xy + k^2 x \implies k = 34$. 2. Injectivity & Discontinuity Backtracking: - Suppose $f(y_1) = f(y_2)$. - $34 x f(y_1) + f(f(x)) = 34 x f(y_2) + f(f(x)) \implies f(y_1) = f(y_2)$ for $x \neq 0$. Thus...
Assume $f(x) = a x + b$. Then $a(ax(ay+b) + ax+b) + b = 34x(ay+b) + a(ax+b)+b$. Comparing coefficients gives $a=34, b=0$. No other solutions exist because all functions are polynomials.
**Analytic Classification:** 1. **Trivial Case**: $f(x) \equiv 0$. 2. **Injectivity**: If $f(y_1) = f(y_2)$, then for $x \neq 0$, $34 x f(y_1) + f(f(x)) = 34 x f(y_2) + f(f(x)) \implies f$ is injective. 3. **Hamel Elimination**: Multiplicativity $f(x^2) = \frac{1}{34} f(x)^2 \ge 0$ enforces positivity on $\mathbb{R}^+...
Chosen proves strict injectivity and eliminates non-measurable Hamel bases; Rejected assumes polynomial form without justification.
How does boundedness on a set of positive Lebesgue measure force linearity in Cauchy equations?
<thought>Steinhaus Theorem on difference sets A - A containing an interval around 0.</thought> **Resolution**: By Steinhaus Theorem, $A - A$ contains an open neighborhood of 0, ensuring continuity at 0 and linearity everywhere.
math_r1_00026
Higher Algebra & Functional Equations
Higher Algebra
AIME / USAMO
Competition Benchmark: IMO Shortlist Benchmark [Problem #ALG-VIETA-01025] Domain: Higher Algebra & Polynomials Problem Statement: Let $x_1, x_2, x_3$ be roots of $P(t) = t^3 - 8t^2 + 15t - 9 = 0$. Using Newton's sums $p_k = x_1^k + x_2^k + x_3^k$, compute $p_4$ exactly.
<thought> 1. Newton's Sums Recurrence: - $p_1 = e_1 = 8$. - $p_2 = e_1 p_1 - 2 e_2 = 8(8) - 2(15) = 34$. - $p_3 = e_1 p_2 - e_2 p_1 + 3 e_3 = 8(34) - 15(8) + 3(9) = 179$. - $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 8(179) - 15(34) + 9(8) = 994$. </thought>
<thought> 1. Newton's Sums Recurrence: - $p_1 = e_1 = 8$. - $p_2 = e_1 p_1 - 2 e_2 = 8(8) - 2(15) = 34$. - $p_3 = e_1 p_2 - e_2 p_1 + 3 e_3 = 8(34) - 15(8) + 3(9) = 179$. - $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 8(179) - 15(34) + 9(8) = 994$. </thought> **Calculation via Newton-Girard Identities:** 1. $p_1 ...
Compute roots directly using Cardano's formula. Since $p_4 = (x_1+x_2+x_3)^4 = 8^4 = 4096$, we obtain the answer without Newton identities.
**Calculation via Newton-Girard Identities:** 1. $p_1 = e_1 = 8$. 2. $p_2 = e_1 p_1 - 2e_2 = 34$. 3. $p_3 = e_1 p_2 - e_2 p_1 + 3e_3 = 179$. 4. $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 8(179) - 15(34) + 9(8) = \boxed{994}$$
Chosen applies Newton-Girard identities correctly subtracting cross-terms; Rejected equates $\sum x_i^4$ with $(\sum x_i)^4$.
How do Newton sums generalize to arbitrary degree $n$ polynomials?
<thought>Logarithmic derivative of polynomial generating function P'(t)/P(t).</thought> **Resolution**: Expanding $\frac{P'(t)}{P(t)} = \sum_{k=1}^\infty p_k t^{-k}$ generates the general Newton recurrence for all degrees.
math_r1_00027
Higher Algebra & Functional Equations
Higher Algebra
IMO SHORTLIST
Competition Benchmark: Putnam Mathematical Competition [Problem #ALG-JENSEN-01026] Domain: Higher Algebra & Inequalities Problem Statement: For positive reals $a, b, c$ with $a + b + c = 1$, prove the strict convexity bound: \[ \frac{a}{\sqrt{1 - a}} + \frac{b}{\sqrt{1 - b}} + \frac{c}{\sqrt{1 - c}} \ge \frac{\sqrt{6}}...
<thought> 1. Convexity Verification: - Define $f(x) = \frac{x}{\sqrt{1-x}}$ for $x \in (0, 1)$. - $f'(x) = \frac{1 - x/2}{(1-x)^{3/2}}$. - $f''(x) = \frac{3(1-x/2)}{2(1-x)^{5/2}} - \frac{1/2}{(1-x)^{3/2}} = \frac{4 - x}{4(1-x)^{5/2}} > 0$. - Since $f''(x) > 0$, $f$ is strictly convex on $(0, 1)$. 2. Jensen ...
<thought> 1. Convexity Verification: - Define $f(x) = \frac{x}{\sqrt{1-x}}$ for $x \in (0, 1)$. - $f'(x) = \frac{1 - x/2}{(1-x)^{3/2}}$. - $f''(x) = \frac{3(1-x/2)}{2(1-x)^{5/2}} - \frac{1/2}{(1-x)^{3/2}} = \frac{4 - x}{4(1-x)^{5/2}} > 0$. - Since $f''(x) > 0$, $f$ is strictly convex on $(0, 1)$. 2. Jensen ...
Apply Cauchy-Schwarz directly: $(\sum a/\sqrt{1-a})(\sum a\sqrt{1-a}) \ge (\sum a)^2 = 1$. This gives lower bound 1, which is weaker than $\sqrt{6}/2$.
**Proof via Jensen's Inequality:** 1. **Convexity**: Let $f(x) = \frac{x}{\sqrt{1-x}}$. Second derivative $f''(x) = \frac{4 - x}{4(1-x)^{5/2}} > 0$ on $(0, 1)$, confirming strict convexity. 2. **Jensen's Bound**: \[ \frac{f(a) + f(b) + f(c)}{3} \ge f\left(\frac{a+b+c}{3}\right) = f\left(\frac{1}{3}\right) = \frac{1/3}...
Chosen establishes second-derivative convexity $f''(x) > 0$ and applies Jensen to hit the exact tight bound; Rejected uses a loose Cauchy-Schwarz.
How does Karamata's Majorization theorem generalize Jensen for sorted sequences?
<thought>Majorization x \succ y implies \sum f(x_i) \ge \sum f(y_i) for convex f.</thought> **Resolution**: Karamata proves that if $(x) \succ (y)$ in majorization order, $\sum f(x_i) \ge \sum f(y_i)$ for any convex function $f$.
math_r1_00028
Higher Algebra & Functional Equations
Higher Algebra
PUTNAM / RESEARCH
Competition Benchmark: Putnam Mathematical Competition [Problem #ALG-CHEBY-01027] Domain: Higher Algebra & Polynomials Problem Statement: Derive the closed-form expression for the Chebyshev polynomial of the first kind $T_{7}(x)$ satisfying $T_{7}(\cos \theta) = \cos(7\theta)$. Find all real roots of $T_{7}(x) = 0$ in ...
<thought> 1. Recurrence Relation: - $T_0(x) = 1, T_1(x) = x$. - $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$ from $\cos((n+1)\theta) + \cos((n-1)\theta) = 2\cos\theta\cos(n\theta)$. 2. Exact Roots: - $\cos(n\theta) = 0 \implies n\theta_k = \frac{2k-1}{2}\pi \implies \theta_k = \frac{2k-1}{2n}\pi$ for $k = 1, \dots, n...
<thought> 1. Recurrence Relation: - $T_0(x) = 1, T_1(x) = x$. - $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$ from $\cos((n+1)\theta) + \cos((n-1)\theta) = 2\cos\theta\cos(n\theta)$. 2. Exact Roots: - $\cos(n\theta) = 0 \implies n\theta_k = \frac{2k-1}{2}\pi \implies \theta_k = \frac{2k-1}{2n}\pi$ for $k = 1, \dots, n...
Since $T_n(x)$ is a polynomial of degree 7, its roots are equally spaced numbers between -1 and 1: $x_k = -1 + \frac{2k}{7}$, which ignores the non-linear cosine density.
**Derivation & Spectral Decomposition:** 1. **Recurrence**: $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$. 2. **Trigonometric Closed Form**: $T_{7}(x) = \cos(7 \arccos x)$ for $x \in [-1, 1]$. 3. **Roots**: Setting $\cos(7\theta) = 0$ gives $n$ distinct real roots: \[ \boxed{x_k = \cos\left(\frac{2k - 1}{2 \cdot 7} \pi\right)...
Chosen derives exact non-linear Chebyshev nodes $x_k = \cos(\frac{2k-1}{2n}\pi)$; Rejected assumes a naive uniform arithmetic progression.
Why do Chebyshev nodes minimize the Runge interpolation error phenomenon?
<thought>Equioscillation and minimization of the monic polynomial supremum norm ||P||_\infty = 2^{1-n}.</thought> **Resolution**: Chebyshev roots minimize the product $\prod |x - x_k|$, making the monic polynomial norm $\|T_n\|_\infty = 2^{1-n}$ minimal on $[-1, 1]$.
math_r1_00029
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-AMGM-01028] Domain: Formal Logic & Real Analysis Theorem Statement: Formalize and prove in Lean 4 that for non-negative reals $a, b$, $2\sqrt{ab} \le a + b$.
<thought> 1. Lean 4 Formalization Architecture: - Target Theorem: $\forall (a \, b : \mathbb{R}), 0 \le a \to 0 \le b \to 2 \cdot \sqrt{a \cdot b} \le a + b$. - Algebraic identity in Lean: $(a + b) - 2\sqrt{ab} = (\sqrt{a} - \sqrt{b})^2 \ge 0$. - Tactics: `have`, `calc`, `sq_nonneg`, `Real.sq_sqrt`, `linarith`...
<thought> 1. Lean 4 Formalization Architecture: - Target Theorem: $\forall (a \, b : \mathbb{R}), 0 \le a \to 0 \le b \to 2 \cdot \sqrt{a \cdot b} \le a + b$. - Algebraic identity in Lean: $(a + b) - 2\sqrt{ab} = (\sqrt{a} - \sqrt{b})^2 \ge 0$. - Tactics: `have`, `calc`, `sq_nonneg`, `Real.sq_sqrt`, `linarith`...
```lean4 theorem am_gm_two (a b : ℝ) (ha : 0 ≀ a) (hb : 0 ≀ b) : 2 * Real.sqrt (a * b) ≀ a + b := by have h := (a - b)^2 linarith ```
```lean4 import Mathlib.Data.Real.Basic import Mathlib.Data.Real.Sqrt import Mathlib.Tactic.Positivity import Mathlib.Tactic.Linarith /-- AM-GM Inequality for n = 2 in Lean 4 -/ theorem am_gm_two (a b : ℝ) (ha : 0 ≀ a) (hb : 0 ≀ b) : 2 * Real.sqrt (a * b) ≀ a + b := by have h_sq : 0 ≀ (Real.sqrt a - Real.sqrt b)...
Chosen provides a complete Mathlib4 calc chain with square root transformations; Rejected misuses $(a-b)^2$ which cannot eliminate square roots.
How do you prove strict equality condition $2\sqrt{ab} = a + b \iff a = b$?
<thought>By sq_eq_zero_iff and Real.sqrt_inj.</thought> **Resolution**: Equality holds when $(\sqrt{a} - \sqrt{b})^2 = 0 \iff \sqrt{a} = \sqrt{b} \iff a = b$.
math_r1_00030
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-BERNOULLI-01029] Domain: Formal Logic & Analysis Theorem Statement: Formalize and prove Bernoulli's Inequality in Lean 4 for all $n : \mathbb{N}$ and $x \ge -1$: \[ (1 + x)^n \ge 1 + n x \]
<thought> 1. Mathematical Induction in Lean 4: - Base case n = 0: (1+x)^0 = 1 \ge 1 + 0 = 1. Trivial by ring / norm_num. - Inductive step: Assume (1+x)^k \ge 1 + kx. - Multiply by (1+x) \ge 0: (1+x)^{k+1} \ge (1 + kx)(1 + x) = 1 + (k+1)x + k x^2 \ge 1 + (k+1)x. - Tactics: induction', linarith, nlinarith, mu...
<thought> 1. Mathematical Induction in Lean 4: - Base case n = 0: (1+x)^0 = 1 \ge 1 + 0 = 1. Trivial by ring / norm_num. - Inductive step: Assume (1+x)^k \ge 1 + kx. - Multiply by (1+x) \ge 0: (1+x)^{k+1} \ge (1 + kx)(1 + x) = 1 + (k+1)x + k x^2 \ge 1 + (k+1)x. - Tactics: induction', linarith, nlinarith, mu...
```lean4 theorem bernoulli_inequality (x : ℝ) (n : β„•) : 1 + (n : ℝ) * x ≀ (1 + x) ^ n := by induction' n with k hk Β· simp Β· nlinarith ```
```lean4 import Mathlib.Data.Real.Basic import Mathlib.Tactic.Linarith import Mathlib.Tactic.Ring /-- Bernoulli's Inequality by induction in Lean 4 -/ theorem bernoulli_inequality (x : ℝ) (hx : -1 ≀ x) (n : β„•) : 1 + (n : ℝ) * x ≀ (1 + x) ^ n := by induction' n with k hk Β· simp Β· have h_pos : 0 ≀ 1 + x := by ...
Chosen establishes $0 \le 1+x$ positivity and proves inductive step via calc; Rejected omits hypothesis $x \ge -1$ making the theorem false for negative odd powers.
Why is the hypothesis $x \ge -1$ essential in the Lean 4 proof?
<thought>Multiplying inequalities by a negative term reverses the inequality sign.</thought> **Resolution**: If $1+x < 0$, `mul_le_mul_of_nonneg_right` fails because multiplication by a negative value reverses $\le$ to $\ge$.
math_r1_00031
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-SQRT2-01030] Domain: Formal Logic & Number Theory Theorem Statement: Formalize and prove in Lean 4 that $\sqrt{2}$ is irrational by showing there exist no coprime natural numbers $p, q$ with $p^2 = 2q^2$.
<thought> 1. Classical Parity Proof in Lean 4: - Target: `theorem sqrt_two_irrational (p q : β„•) (hq : q β‰  0) (h_coprime : Nat.Coprime p q) : p ^ 2 β‰  2 * q ^ 2`. - If $p^2 = 2q^2$, then $2 \mid p^2 \implies 2 \mid p$. - Write $p = 2k$, then $4k^2 = 2q^2 \implies q^2 = 2k^2 \implies 2 \mid q$. - Contradiction...
<thought> 1. Classical Parity Proof in Lean 4: - Target: `theorem sqrt_two_irrational (p q : β„•) (hq : q β‰  0) (h_coprime : Nat.Coprime p q) : p ^ 2 β‰  2 * q ^ 2`. - If $p^2 = 2q^2$, then $2 \mid p^2 \implies 2 \mid p$. - Write $p = 2k$, then $4k^2 = 2q^2 \implies q^2 = 2k^2 \implies 2 \mid q$. - Contradiction...
```lean4 theorem sqrt_two_irrational (p q : β„•) : p ^ 2 β‰  2 * q ^ 2 := by intro h have h2 : p = 0 := by linarith sorry ```
```lean4 import Mathlib.Data.Nat.Prime import Mathlib.Tactic.Ring /-- Proof of irrationality of sqrt(2) in Lean 4 -/ theorem sqrt_two_irrational (p q : β„•) (h_coprime : p.Coprime q) : p ^ 2 β‰  2 * q ^ 2 := by intro h_eq have h_even_p2 : 2 ∣ p ^ 2 := by use q ^ 2 exact h_eq.symm have h_even_p : 2 ∣ p :=...
Chosen uses formal Mathlib4 `Nat.Prime.dvd_of_dvd_pow` to prove parity contradiction on coprime gcd; Rejected collapses to sorry.
How does this generalize to $\sqrt{n}$ for any non-square integer $n$?
<thought>By p-adic valuation v_p(p^2) = 2 v_p(p) is always even, while v_p(n q^2) is odd for prime p with odd exponent in n.</thought> **Resolution**: Fundamental Theorem of Arithmetic implies $v_p(p^2)$ is even, whereas $v_p(n q^2) = v_p(n) + 2v_p(q)$ is odd, yielding an immediate contradiction.
math_r1_00032
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-LAGRANGE-01031] Domain: Formal Logic & Abstract Algebra Theorem Statement: Formalize in Lean 4 the coset partition equality: for any finite group $G$ and subgroup $H \le G$, $|G| = [G : H] \cdot |H|$.
<thought> 1. Lean 4 Abstract Algebra Architecture: - Typeclasses: `Group G`, `Fintype G`, `Subgroup H`. - Left coset equivalence relation $a \sim b \iff a^{-1}b \in H$. - Bijections between distinct cosets: $x H \cong H$ via $h \mapsto x h$. </thought>
<thought> 1. Lean 4 Abstract Algebra Architecture: - Typeclasses: `Group G`, `Fintype G`, `Subgroup H`. - Left coset equivalence relation $a \sim b \iff a^{-1}b \in H$. - Bijections between distinct cosets: $x H \cong H$ via $h \mapsto x h$. </thought> ```lean4 import Mathlib.Algebra.Group.Subgroup.Basic impo...
```lean4 theorem lagrange_group_order {G : Type*} [Group G] (H : Subgroup G) : Fintype.card G = Fintype.card H := by rfl ```
```lean4 import Mathlib.Algebra.Group.Subgroup.Basic import Mathlib.GroupTheory.Index import Mathlib.Data.Fintype.Card /-- Lagrange's Theorem for Finite Groups in Lean 4 -/ theorem lagrange_group_order {G : Type*} [Group G] [Fintype G] (H : Subgroup G) [Fintype H] : Fintype.card G = H.index * Fintype.card H := by ...
Chosen applies correct Mathlib4 subgroup quotient cardinality theorem; Rejected falsely claims subgroup order equals group order by rfl.
What is the corollary for prime-order groups?
<thought>If |G| = p, proper subgroups have order 1 or p, forcing G to be cyclic.</thought> **Resolution**: Every group of prime order $p$ has no non-trivial proper subgroups, and is therefore cyclic ($G \cong \mathbb{Z}/p\mathbb{Z}$).
math_r1_00033
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
PUTNAM / RESEARCH
Competition Benchmark: Lean 4 Mathlib Research [Problem #COMB-RAMSEY-01032] Domain: Advanced Combinatorics Problem Statement: In complete graph $K_n$, prove using the probabilistic method with ErdΕ‘s expectation and the Symmetric LovΓ‘sz Local Lemma (LLL) that diagonal Ramsey number satisfies $R(7, 7) > \frac{\sqrt{2}7}{...
<thought> 1. Random Edge Coloring: - Each edge colored Red/Blue with $p = 1/2$. - Probability of monochromatic $K_k$: $P(A_S) = 2^{1 - \binom{k}{2}}$. 2. Dependency Graph & LLL: - Two subsets $S, T$ are dependent iff $|S \cap T| \ge 2$. - Degree $d = \binom{k}{2} \binom{n-2}{k-2}$. - LLL Condition: $e \c...
<thought> 1. Random Edge Coloring: - Each edge colored Red/Blue with $p = 1/2$. - Probability of monochromatic $K_k$: $P(A_S) = 2^{1 - \binom{k}{2}}$. 2. Dependency Graph & LLL: - Two subsets $S, T$ are dependent iff $|S \cap T| \ge 2$. - Degree $d = \binom{k}{2} \binom{n-2}{k-2}$. - LLL Condition: $e \c...
Color alternating edges Red and Blue. In any graph, there are at most $n/2$ edges of one color, so no monochromatic clique of size 7 can exist for $n < 2^7$.
**Probabilistic Bound via Symmetric LovΓ‘sz Local Lemma:** 1. **Bad Event Probability**: For subset $S$ of size $7$, $P(A_S) = 2^{1 - \binom{7}{2}}$. 2. **Dependency Degree**: $d = \binom{7}{2} \binom{n-2}{7-2} < \frac{7^2}{2} \frac{n^{7-2}}{(7-2)!}$. 3. **LLL Criterion**: $e \cdot P(A_S) \cdot (d + 1) \le 1$ ensures $...
Chosen derives the exact dependency degree $d$ and invokes Symmetric LLL; Rejected makes an invalid deterministic alternating coloring claim.
How does Spencer's polynomial method compare with LLL?
<thought>Spencer's fractional coloring achieves optimal constants within the probabilistic method.</thought> **Resolution**: Spencer's theorem matches the asymptotic constant while providing an algorithmic Moser-Tardos resample algorithm.
math_r1_00034
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
AIME / USAMO
Competition Benchmark: Lean 4 Mathlib Research [Problem #COMB-DERANGE-01033] Domain: Advanced Combinatorics Problem Statement: Using the Principle of Inclusion-Exclusion, derive the exact closed formula for derangements $D_n = !4$ (permutations of $4$ elements with no fixed points) and evaluate $D_{4}$ exactly.
<thought> 1. Inclusion-Exclusion Formulation: - Let $A_i$ be permutations fixing $i$. $|A_i| = (n-1)!$. - Intersection of $k$ sets: $|igcap_{j=1}^k A_{i_j}| = (n-k)!$. - $D_n = n! - \sum |A_i| + \sum |A_i \cap A_j| - \dots = n! \sum_{k=0}^n \frac{(-1)^k}{k!}$. 2. Evaluation for n = 4: - $D_{n} = \lfloor \f...
<thought> 1. Inclusion-Exclusion Formulation: - Let $A_i$ be permutations fixing $i$. $|A_i| = (n-1)!$. - Intersection of $k$ sets: $|igcap_{j=1}^k A_{i_j}| = (n-k)!$. - $D_n = n! - \sum |A_i| + \sum |A_i \cap A_j| - \dots = n! \sum_{k=0}^n \frac{(-1)^k}{k!}$. 2. Evaluation for n = 4: - $D_{n} = \lfloor \f...
Since each element has $n-1$ choices for its position, $D_n = (n-1)^n$. For $n=4$, this gives $(4-1)^{4}$, which ignores dependency among cycles.
**Derivation via Inclusion-Exclusion:** 1. **Union of Fixed Points**: $|\bigcup_{i=1}^{4} A_i| = \sum_{k=1}^{4} (-1)^{k-1} \binom{4}{k} (4-k)! = 4! \sum_{k=1}^{4} \frac{(-1)^{k-1}}{k!}$. 2. **Complement**: $D_{4} = 4! \left(1 - \sum_{k=1}^{4} \frac{(-1)^{k-1}}{k!}\right) = 4! \sum_{k=0}^{4} \frac{(-1)^k}{k!}$. \boxed...
Chosen uses formal PIE with alternating sum converging to $n!/e$; Rejected uses independent choices $(n-1)^n$ allowing duplicate assignments.
How does the recurrence $D_n = (n-1)(D_{n-1} + D_{n-2})$ relate to permutation cycle decomposition?
<thought>Element 1 swaps with element k (D_{n-2}) or is part of a larger cycle (D_{n-1}).</thought> **Resolution**: Partitioning based on whether element 1 belongs to a 2-cycle or cycle of length $\ge 3$ yields $D_n = (n-1)(D_{n-1} + D_{n-2})$.
math_r1_00035
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
PUTNAM / IMO HARD
Competition Benchmark: USAMO National Olympiad [Problem #COMB-CAYLEY-01034] Domain: Advanced Combinatorics Problem Statement: Prove Cayley's Formula for the number of labeled trees on $7$ vertices $T_{7} = 7^{7-2}$ by establishing a bijective correspondence with PrΓΌfer sequences of length $7-2$.
<thought> 1. PrΓΌfer Sequence Bijection: - Tree \to PrΓΌfer: Successively remove leaf with smallest label and record its neighbor. Repeat until 2 vertices remain. - Result is a sequence of length n - 2 with entries in \{1, \dots, n\}. - Number of such sequences is n^{n-2}. - Inverse algorithm reconstructs uni...
<thought> 1. PrΓΌfer Sequence Bijection: - Tree \to PrΓΌfer: Successively remove leaf with smallest label and record its neighbor. Repeat until 2 vertices remain. - Result is a sequence of length n - 2 with entries in \{1, \dots, n\}. - Number of such sequences is n^{n-2}. - Inverse algorithm reconstructs uni...
A tree has $n-1$ edges chosen from $\binom{n}{2}$ possible edges. So $T_n = \binom{\binom{n}{2}}{n-1}$, which counts all acyclic and cyclic subgraphs indiscriminately.
**Bijective Proof via PrΓΌfer Sequences:** 1. **Tree to Sequence**: Iteratively delete the minimal leaf $v$ and append its unique neighbor $u$ to sequence $P$. The process terminates at length $7-2$. 2. **Reconstruction**: Each vertex appears in $P$ exactly $\text{deg}(v) - 1$ times. The minimal missing label is the fi...
Chosen establishes explicit bijection with PrΓΌfer sequences; Rejected naively counts arbitrary subsets of edges $\binom{N}{n-1}$.
How does Matrix-Tree Theorem prove Cayley's formula algebraically?
<thought>Laplacian matrix of K_n has eigenvalues n with multiplicity n-1, so det cofactor is n^{n-2}.</thought> **Resolution**: The Laplacian of $K_n$ has spectrum $\{0, n, \dots, n\}$, and any cofactor equals $\frac{1}{n} \prod_{i=1}^{n-1} n = n^{n-2}$.
math_r1_00036
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
AIME / USAMO
Competition Benchmark: AIME 2026 Invitational [Problem #COMB-CATALAN-01035] Domain: Advanced Combinatorics Problem Statement: Using Ordinary Generating Functions (OGF), solve the Catalan recurrence $C_0 = 1, C_{n+1} = \sum_{i=0}^n C_i C_{n-i}$ and compute $C_{9}$.
<thought> 1. OGF Setup: - Let C(x) = \sum_{n=0}^\infty C_n x^n. - C(x) = 1 + x C(x)^2 \implies x C(x)^2 - C(x) + 1 = 0. - Roots: C(x) = \frac{1 \pm \sqrt{1 - 4x}}{2x}. - Since C(0) = 1, we select minus sign: C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}. 2. Taylor Expansion: - \sqrt{1-4x} = 1 - \sum_{n=1}^\infty \...
<thought> 1. OGF Setup: - Let C(x) = \sum_{n=0}^\infty C_n x^n. - C(x) = 1 + x C(x)^2 \implies x C(x)^2 - C(x) + 1 = 0. - Roots: C(x) = \frac{1 \pm \sqrt{1 - 4x}}{2x}. - Since C(0) = 1, we select minus sign: C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}. 2. Taylor Expansion: - \sqrt{1-4x} = 1 - \sum_{n=1}^\infty \...
The recurrence $C_{n+1} = \sum C_i C_{n-i}$ is solved by $C_n = 2^n$ because at each step we make a binary choice.
**Closed Form via Algebraic Generating Function:** 1. **Functional Equation**: $C(x) = 1 + x C(x)^2 \implies C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}$. 2. **Binomial Series**: $\sqrt{1 - 4x} = 1 - 2 \sum_{n=1}^\infty \frac{1}{n} \binom{2n-2}{n-1} x^n$. 3. **Extraction**: $C_n = \frac{1}{n+1} \binom{2n}{n}$. \boxed{C_{9} =...
Chosen derives algebraic OGF $C(x) = (1-\sqrt{1-4x})/2x$ and extracts exact coefficients; Rejected mistakes Catalan growth for exponential $2^n$.
How does Dyck path reflection principle prove $C_n = \binom{2n}{n} - \binom{2n}{n+1}$?
<thought>Reflecting bad paths that touch y = -1 across the line y = -1 establishes a bijection to paths ending at (2n, -2).</thought> **Resolution**: AndrΓ©'s Reflection Principle reflects the portion of an invalid path after its first touch of $y=-1$, mapping invalid paths bijectively to $\binom{2n}{n-1}$.
math_r1_00037
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: AIME 2026 Invitational [Problem #GEOM-EULER-01036] Domain: Euclidean Geometry Problem Statement: In triangle $\triangle ABC$, inradius is $r = 48$ and circumradius is $R = 130$. Prove Euler's formula $d^2 = R(R - 2r)$ for distance $d = IO$ and compute $d$.
<thought> 1. Geometric Invariant: - Power of incenter $I$ with respect to circumcircle: $R^2 - d^2 = AI \cdot IM$. - Trillium Theorem: $IM = MB$. - $AI = r / \sin(A/2)$, $MB = 2R \sin(A/2)$. - $AI \cdot MB = 2Rr \implies d^2 = R(R - 2r) = 4420$. </thought>
<thought> 1. Geometric Invariant: - Power of incenter $I$ with respect to circumcircle: $R^2 - d^2 = AI \cdot IM$. - Trillium Theorem: $IM = MB$. - $AI = r / \sin(A/2)$, $MB = 2R \sin(A/2)$. - $AI \cdot MB = 2Rr \implies d^2 = R(R - 2r) = 4420$. </thought> **Synthetic Proof via Trillium Lemma:** 1. **Powe...
By Pythagoras theorem in $\triangle AIO$, $d^2 = R^2 + r^2$. For $R=130, r=48$, $d = \sqrt{19204}$, which ignores non-orthogonality.
**Synthetic Proof via Trillium Lemma:** 1. **Power of Incenter**: $R^2 - d^2 = AI \cdot IM$. 2. **Trillium Lemma**: $IM = MB = 2R \sin(A/2)$ and $AI = \frac{r}{\sin(A/2)}$. 3. **Product**: $AI \cdot IM = 2Rr \implies \boxed{d^2 = R(R - 2r) = 4420 \implies d = \sqrt{4420}}$.
Chosen applies Trillium Lemma and Power of Point on circumcircle; Rejected falsely assumes $OI \perp AI$ applying Pythagorean theorem.
How does $d^2 \ge 0$ immediately prove Euler's inequality $R \ge 2r$?
<thought>Real Euclidean distance requires d^2 >= 0, so R(R-2r) >= 0.</thought> **Resolution**: Since $R > 0$, $R(R-2r) = d^2 \ge 0$ yields $R \ge 2r$, with equality $R = 2r$ if and only if $d = 0$ (equilateral triangle).
math_r1_00038
Euclidean & Projective Geometry
Euclidean
USAMO / IMO
Competition Benchmark: USAMO National Olympiad [Problem #GEOM-PTOLEMY-01037] Domain: Euclidean & Projective Geometry Problem Statement: In cyclic quadrilateral $ABCD$ with sides $AB=9, BC=10, CD=5, DA=10$, prove Ptolemy's Theorem $AC \cdot BD = AB \cdot CD + BC \cdot DA$ using circular inversion.
<thought> 1. Geometric Inversion Method: - Invert with center A and arbitrary radius R. - Circle passing through A, B, C, D maps to a straight line containing B', C', D' in order. - Segment lengths under inversion: X'Y' = \frac{R^2 \cdot XY}{AX \cdot AY}. - Collinear segments: B'D' = B'C' + C'D'. - Subst...
<thought> 1. Geometric Inversion Method: - Invert with center A and arbitrary radius R. - Circle passing through A, B, C, D maps to a straight line containing B', C', D' in order. - Segment lengths under inversion: X'Y' = \frac{R^2 \cdot XY}{AX \cdot AY}. - Collinear segments: B'D' = B'C' + C'D'. - Subst...
Apply law of cosines to $\triangle ABC$ and $\triangle ADC$. Then add the two diagonal formulas directly without matching opposite supplementary angles.
**Proof via Circular Inversion:** 1. **Inversion Action**: Invert the figure at center $A$. The circumcircle through $A$ maps to a straight line on which points $B', C', D'$ lie in collinear order. 2. **Metric Transformation**: $B'D' = \frac{R^2 \cdot BD}{AB \cdot AD}$, $B'C' = \frac{R^2 \cdot BC}{AB \cdot AC}$, $C'D'...
Chosen establishes rigorous proof via circular inversion mapping circles to lines; Rejected fails to eliminate cosines of opposite angles.
How does Ptolemy's inequality hold for non-cyclic quadrilaterals?
<thought>Triangle inequality on the inverted points B', C', D' gives B'D' <= B'C' + C'D'.</thought> **Resolution**: For non-cyclic quadrilaterals, $B', C', D'$ form a non-degenerate triangle, so triangle inequality yields strict $AC \cdot BD < AB \cdot CD + BC \cdot DA$.
math_r1_00039
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: AIME 2026 Invitational [Problem #GEOM-RADICAL-01038] Domain: Euclidean Geometry Problem Statement: Let $\omega_1, \omega_2, \omega_3$ be three non-concentric circles whose centers are not collinear. Prove that the three radical axes pairwise intersect at a unique point $P$ (the Radical Center).
<thought> 1. Power of a Point & Radical Axis Definition: - For circle \omega(O, R), \text{Pow}(X, \omega) = d(X, O)^2 - R^2. - Radical axis L_{12} = \{X : \text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)\}. 2. Intersection Proof: - Let P = L_{12} \cap L_{23}. - Then \text{Pow}(P, \omega_1) = \text{Pow}(P, ...
<thought> 1. Power of a Point & Radical Axis Definition: - For circle \omega(O, R), \text{Pow}(X, \omega) = d(X, O)^2 - R^2. - Radical axis L_{12} = \{X : \text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)\}. 2. Intersection Proof: - Let P = L_{12} \cap L_{23}. - Then \text{Pow}(P, \omega_1) = \text{Pow}(P, ...
Since all circles are round, their chords intersect at the center of mass of the three centers.
**Proof of Radical Center Concurrence:** 1. **Linearity of Radical Axis**: The locus $\text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)$ is a straight line perpendicular to the line of centers $O_1 O_2$. 2. **Intersection**: Since centers are non-collinear, line $L_{12}$ and $L_{23}$ intersect at unique point $P$. 3. ...
Chosen applies analytic Power of Point transitivity $\text{Pow}(P, \omega_1) = \text{Pow}(P, \omega_3)$; Rejected makes a nonsensical center-of-mass heuristic.
What happens if the centers $O_1, O_2, O_3$ are collinear?
<thought>The radical axes are parallel lines perpendicular to the line of centers.</thought> **Resolution**: If centers are collinear, the three radical axes are mutually parallel perpendiculars (intersecting at infinity in projective plane).
math_r1_00040
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: Putnam Mathematical Competition [Problem #GEOM-SIMSON-01039] Domain: Euclidean & Projective Geometry Problem Statement: Prove the Simson Line Theorem: The orthogonal projections $P_a, P_b, P_c$ of point $P$ onto sides $BC, CA, AB$ of $\triangle ABC$ are collinear if and only if $P$ lies on the ci...
<thought> 1. Cyclic Quadrilateral Angle Chasing: - Points P, P_a, B, P_c are concyclic on circle with diameter PB. - Points P, P_a, C, P_b are concyclic on circle with diameter PC. - Collinearity of P_a, P_b, P_c requires \angle P P_a P_c = \angle P P_a P_b. - In cyclic quads: \angle P P_a P_c = \angle P B ...
<thought> 1. Cyclic Quadrilateral Angle Chasing: - Points P, P_a, B, P_c are concyclic on circle with diameter PB. - Points P, P_a, C, P_b are concyclic on circle with diameter PC. - Collinearity of P_a, P_b, P_c requires \angle P P_a P_c = \angle P P_a P_b. - In cyclic quads: \angle P P_a P_c = \angle P B ...
Connect the three pedal points to form $\triangle P_a P_b P_c$. Use Heron's formula for area. Setting area to 0 gives $P_a = P_b$, which requires $P$ to be at a vertex.
**Angle-Chasing Proof of Simson's Theorem:** 1. **Concyclic Pedal Quadrilaterals**: Since $\angle P P_a B = \angle P P_c B = 90^\circ$, quadrilateral $P P_a B P_c$ is cyclic with diameter $PB$. Likewise, $P P_a C P_b$ is cyclic with diameter $PC$. 2. **Inscribed Angle Equivalence**: $\angle P P_a P_c = \angle P B A$ a...
Chosen uses cyclic pedal circles to prove $\angle PBA = \angle PCA$ iff $P$ on circumcircle; Rejected confuses collinearity with point collapse $P_a = P_b$.
How does the Simson line bisect the segment $P H$ where $H$ is the orthocenter?
<thought>Homothety of ratio 1/2 from H maps circumcircle to nine-point circle, placing midpoint of PH on Simson line.</thought> **Resolution**: The midpoint of $PH$ lies on the nine-point circle and the Simson line of $P$, meaning the Simson line bisects $PH$.
math_r1_00041
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
AIME / IMO HARD
Competition Benchmark: IMO Shortlist Benchmark [Problem #NT-PELL-01040] Domain: Olympiad Number Theory Problem Statement: Find all positive integer pairs $(x, y)$ such that $x^2 - 46y^2 = 1$ with $x \equiv 142 \pmod{157}$. Prove that there are infinitely many such pairs and determine the periodicity of solutions in $\m...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{46}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 157 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(46\tilde{y}_1^2)$ is a quadratic non...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{46}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 157 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(46\tilde{y}_1^2)$ is a quadratic non...
Substitute $x=142$ directly into $x^2 - 46y^2 = 1$. Then $y = \sqrt{(142^2-1)/46}$. Since $142^2-1$ can be made divisible by $46$, solutions exist without checking group periodicity.
**Theorem & Complete Algebraic Number Theory Solution:** 1. **Pell Group Action**: Solutions satisfy $x_n + y_n\sqrt{46} = (x_1 + y_1\sqrt{46})^n$. 2. **Periodicity**: Over $\mathbb{F}_{157}$, if $(46/157) = 1$, period divides $p-1$; if $(46/157) = -1$, period divides $p+1$. 3. **Infinitude**: The residue $x_{n_0} \eq...
Chosen proves periodicity in algebraic number field and analyzes quadratic character; Rejected naively evaluates a single value without proving infinitude or periodicity.
How does the norm map $N(\alpha) = 1$ in $\mathbb{F}_{159}$ force the period to divide $p+1$?
<thought>Norm map kernel has order $(p^2-1)/(p-1) = p+1$.</thought> **Lemma**: The cyclic kernel of $N: \mathbb{F}_{159}^\times \to \mathbb{F}_{157}^\times$ has exact cardinality $p+1$.
math_r1_00042
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
IMO SHORTLIST
Competition Benchmark: IMO Shortlist Benchmark [Problem #NT-LTE-01041] Domain: Olympiad Number Theory Problem Statement: Let $p = 13$ be an odd prime. Suppose positive integers $a, b$ satisfy $p \mid (a - b)$ but $p \nmid a$ and $p \nmid b$. Prove the Lifting The Exponent (LTE) identity for $n = 11$: \[ v_{13}(a^{11} -...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
By expanding $a^n - b^n = (a-b)(a^{n-1} + \dots + b^{n-1})$, each of the $n$ terms is equal to $a^{n-1} \pmod p$, so the sum is $n a^{n-1}$. Thus $v_p(a^n - b^n) = v_p(a-b) + v_p(n a^{n-1}) = v_p(a-b) + v_p(n) + (n-1)v_p(a)$, which gives an extra term.
**Proof via Binomial p-adic Expansion:** 1. Let $v_p(a - b) = k \ge 1$. Write $a = b + c$ with $v_p(c) = k$. 2. Expanding $a^n = (b + c)^n = b^n + n b^{n-1} c + \sum_{j=2}^n \binom{n}{j} b^{n-j} c^j$. 3. For $j \ge 2$, $v_p(\binom{n}{j} c^j) \ge j k - v_p(j) > k + v_p(n)$ for odd prime $p$. 4. Therefore $v_p(a^n - b^n...
Chosen establishes p-adic valuations with strict inequality $j k - v_p(j) > k + v_p(n)$; Rejected incorrectly asserts $v_p(a) > 0$ when $p \nmid a$.
Why does LTE fail for $p = 2$ when $n$ is even?
<thought>For p=2 and j=2, 2k - 1 = k + (k-1) equals k when k=1, causing cancellation.</thought> **Resolution**: For $p=2$, the second-order binomial coefficient has $v_2(\binom{n}{2} c^2) = v_2(n) - 1 + 2k$, requiring the extra correction term $v_2(a+b)-1$.
math_r1_00043
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
PUTNAM / IMO HARD
Competition Benchmark: Lean 4 Mathlib Research [Problem #NT-GAUSS-01042] Domain: Olympiad Number Theory Problem Statement: Evaluate the Legendre symbol $\left(\frac{89}{79}\right)$ using Gauss's Law of Quadratic Reciprocity. Prove whether the congruence $x^2 \equiv 89 \pmod{79}$ is solvable in integers.
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 89 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $79 \pmod{89} = 79$. - Decompose into prime factors and apply Euler's criterion. </thou...
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 89 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $79 \pmod{89} = 79$. - Decompose into prime factors and apply Euler's criterion. </thou...
Since both 89 and 79 are odd primes, $(p/q) = -(q/p)$ always holds by Gauss's theorem. Evaluating directly gives the opposite sign.
**Evaluation via Quadratic Reciprocity:** 1. **Parity Check**: Since $p = 89 \equiv 1 \pmod 4$, $(-1)^{\frac{89-1}{2}\frac{79-1}{2}} = 1$. 2. **Reciprocity Application**: $\left(\frac{89}{79}\right) = \left(\frac{79}{89}\right)$. 3. **Modular Reduction**: Evaluating $\left(\frac{79}{89}\right) = 1$. Therefore, the co...
Chosen correctly analyzes $p \equiv 1 \pmod 4$ to obtain $(p/q) = +(q/p)$; Rejected misremembers Quadratic Reciprocity as always having a negative sign.
How does Gauss's Lemma with half-intervals prove quadratic reciprocity?
<thought>Counting lattice points in rectangle of size (p-1)/2 by (q-1)/2.</thought> **Resolution**: Eisenstein's geometric proof counts interior integer points in the rectangle $[1, (p-1)/2] \times [1, (q-1)/2]$.
math_r1_00044
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
AIME / USAMO
Competition Benchmark: Putnam Mathematical Competition [Problem #NT-CRT-01043] Domain: Olympiad Number Theory Problem Statement: Solve the simultaneous system of modular polynomial congruences: \[ x \equiv 16 \pmod{17}, \quad x \equiv 13 \pmod{19} \] and lift the unique solution modulo $M = 323$ to modulo $M^2 = 104329...
<thought> 1. CRT Inversion: - $M = 17 \times 19 = 323$. - $M_1 = 19, M_2 = 17$. - $y_1 = 17^{-1} \pmod{19} = 9$, $y_2 = 19^{-1} \pmod{17} = 9$. - $x_0 = (16 \cdot 19 \cdot 9 + 13 \cdot 17 \cdot 9) \pmod{323} = 203$. 2. Hensel Lifting Step: - For $f(x) = x - x_0 \equiv 0 \pmod M$, derivative $f'(x_0) = 1 ...
<thought> 1. CRT Inversion: - $M = 17 \times 19 = 323$. - $M_1 = 19, M_2 = 17$. - $y_1 = 17^{-1} \pmod{19} = 9$, $y_2 = 19^{-1} \pmod{17} = 9$. - $x_0 = (16 \cdot 19 \cdot 9 + 13 \cdot 17 \cdot 9) \pmod{323} = 203$. 2. Hensel Lifting Step: - For $f(x) = x - x_0 \equiv 0 \pmod M$, derivative $f'(x_0) = 1 ...
Multiply the two equations: $x^2 \equiv 208 \pmod{323}$. Then take square root $x = \sqrt{208}$, which is not integer-preserving.
**Exact CRT & Hensel Lifting:** 1. **BΓ©zout Multipliers**: $19 \cdot 9 \equiv 1 \pmod{17}$ and $17 \cdot 9 \equiv 1 \pmod{19}$. 2. **CRT Base Solution**: $x_0 \equiv 16(19)(9) + 13(17)(9) \equiv 203 \pmod{323}$. 3. **Hensel Lift**: Since $\gcd(f'(x_0), M) = 1$, the unique lift modulo $104329$ is \boxed{x \equiv 203 \p...
Chosen applies rigorous BΓ©zout modular inverses in CRT and verifies Hensel nonsingularity condition; Rejected commits an illegal non-linear multiplication of congruences.
What happens during Hensel lifting if $f'(x_0) \equiv 0 \pmod p$?
<thought>Branching into no solutions or p distinct solutions depending on f(x0) mod p^2.</thought> **Resolution**: If $f'(x_0) \equiv 0 \pmod p$, the lift exists if and only if $f(x_0) \equiv 0 \pmod{p^2}$, generating $p$ distinct lifts.
math_r1_00045
Higher Algebra & Functional Equations
Higher Algebra
PUTNAM / IMO HARD
Competition Benchmark: IMO Shortlist Benchmark [Problem #ALG-CAUCHY-01044] Domain: Higher Algebra & Functional Equations Problem Statement: Find all functions $f: \mathbb{R} \to \mathbb{R}$ satisfying $f(x f(y) + f(x)) = 14 x f(y) + f(f(x))$ for all $x, y \in \mathbb{R}$.
<thought> 1. Trivial and Linear Ansatz: - $f(x) \equiv 0$ is valid. - For $f(x) = k x$, $k^2 xy + k^2 x = 14k xy + k^2 x \implies k = 14$. 2. Injectivity & Discontinuity Backtracking: - Suppose $f(y_1) = f(y_2)$. - $14 x f(y_1) + f(f(x)) = 14 x f(y_2) + f(f(x)) \implies f(y_1) = f(y_2)$ for $x \neq 0$. Thus...
<thought> 1. Trivial and Linear Ansatz: - $f(x) \equiv 0$ is valid. - For $f(x) = k x$, $k^2 xy + k^2 x = 14k xy + k^2 x \implies k = 14$. 2. Injectivity & Discontinuity Backtracking: - Suppose $f(y_1) = f(y_2)$. - $14 x f(y_1) + f(f(x)) = 14 x f(y_2) + f(f(x)) \implies f(y_1) = f(y_2)$ for $x \neq 0$. Thus...
Assume $f(x) = a x + b$. Then $a(ax(ay+b) + ax+b) + b = 14x(ay+b) + a(ax+b)+b$. Comparing coefficients gives $a=14, b=0$. No other solutions exist because all functions are polynomials.
**Analytic Classification:** 1. **Trivial Case**: $f(x) \equiv 0$. 2. **Injectivity**: If $f(y_1) = f(y_2)$, then for $x \neq 0$, $14 x f(y_1) + f(f(x)) = 14 x f(y_2) + f(f(x)) \implies f$ is injective. 3. **Hamel Elimination**: Multiplicativity $f(x^2) = \frac{1}{14} f(x)^2 \ge 0$ enforces positivity on $\mathbb{R}^+...
Chosen proves strict injectivity and eliminates non-measurable Hamel bases; Rejected assumes polynomial form without justification.
How does boundedness on a set of positive Lebesgue measure force linearity in Cauchy equations?
<thought>Steinhaus Theorem on difference sets A - A containing an interval around 0.</thought> **Resolution**: By Steinhaus Theorem, $A - A$ contains an open neighborhood of 0, ensuring continuity at 0 and linearity everywhere.
math_r1_00046
Higher Algebra & Functional Equations
Higher Algebra
AIME / USAMO
Competition Benchmark: Lean 4 Mathlib Research [Problem #ALG-VIETA-01045] Domain: Higher Algebra & Polynomials Problem Statement: Let $x_1, x_2, x_3$ be roots of $P(t) = t^3 - 2t^2 + 1t - 3 = 0$. Using Newton's sums $p_k = x_1^k + x_2^k + x_3^k$, compute $p_4$ exactly.
<thought> 1. Newton's Sums Recurrence: - $p_1 = e_1 = 2$. - $p_2 = e_1 p_1 - 2 e_2 = 2(2) - 2(1) = 2$. - $p_3 = e_1 p_2 - e_2 p_1 + 3 e_3 = 2(2) - 1(2) + 3(3) = 11$. - $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 2(11) - 1(2) + 3(2) = 26$. </thought>
<thought> 1. Newton's Sums Recurrence: - $p_1 = e_1 = 2$. - $p_2 = e_1 p_1 - 2 e_2 = 2(2) - 2(1) = 2$. - $p_3 = e_1 p_2 - e_2 p_1 + 3 e_3 = 2(2) - 1(2) + 3(3) = 11$. - $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 2(11) - 1(2) + 3(2) = 26$. </thought> **Calculation via Newton-Girard Identities:** 1. $p_1 = e_1 = 2...
Compute roots directly using Cardano's formula. Since $p_4 = (x_1+x_2+x_3)^4 = 2^4 = 16$, we obtain the answer without Newton identities.
**Calculation via Newton-Girard Identities:** 1. $p_1 = e_1 = 2$. 2. $p_2 = e_1 p_1 - 2e_2 = 2$. 3. $p_3 = e_1 p_2 - e_2 p_1 + 3e_3 = 11$. 4. $p_4 = e_1 p_3 - e_2 p_2 + e_3 p_1 = 2(11) - 1(2) + 3(2) = \boxed{26}$$
Chosen applies Newton-Girard identities correctly subtracting cross-terms; Rejected equates $\sum x_i^4$ with $(\sum x_i)^4$.
How do Newton sums generalize to arbitrary degree $n$ polynomials?
<thought>Logarithmic derivative of polynomial generating function P'(t)/P(t).</thought> **Resolution**: Expanding $\frac{P'(t)}{P(t)} = \sum_{k=1}^\infty p_k t^{-k}$ generates the general Newton recurrence for all degrees.
math_r1_00047
Higher Algebra & Functional Equations
Higher Algebra
IMO SHORTLIST
Competition Benchmark: Putnam Mathematical Competition [Problem #ALG-JENSEN-01046] Domain: Higher Algebra & Inequalities Problem Statement: For positive reals $a, b, c$ with $a + b + c = 1$, prove the strict convexity bound: \[ \frac{a}{\sqrt{1 - a}} + \frac{b}{\sqrt{1 - b}} + \frac{c}{\sqrt{1 - c}} \ge \frac{\sqrt{6}}...
<thought> 1. Convexity Verification: - Define $f(x) = \frac{x}{\sqrt{1-x}}$ for $x \in (0, 1)$. - $f'(x) = \frac{1 - x/2}{(1-x)^{3/2}}$. - $f''(x) = \frac{3(1-x/2)}{2(1-x)^{5/2}} - \frac{1/2}{(1-x)^{3/2}} = \frac{4 - x}{4(1-x)^{5/2}} > 0$. - Since $f''(x) > 0$, $f$ is strictly convex on $(0, 1)$. 2. Jensen ...
<thought> 1. Convexity Verification: - Define $f(x) = \frac{x}{\sqrt{1-x}}$ for $x \in (0, 1)$. - $f'(x) = \frac{1 - x/2}{(1-x)^{3/2}}$. - $f''(x) = \frac{3(1-x/2)}{2(1-x)^{5/2}} - \frac{1/2}{(1-x)^{3/2}} = \frac{4 - x}{4(1-x)^{5/2}} > 0$. - Since $f''(x) > 0$, $f$ is strictly convex on $(0, 1)$. 2. Jensen ...
Apply Cauchy-Schwarz directly: $(\sum a/\sqrt{1-a})(\sum a\sqrt{1-a}) \ge (\sum a)^2 = 1$. This gives lower bound 1, which is weaker than $\sqrt{6}/2$.
**Proof via Jensen's Inequality:** 1. **Convexity**: Let $f(x) = \frac{x}{\sqrt{1-x}}$. Second derivative $f''(x) = \frac{4 - x}{4(1-x)^{5/2}} > 0$ on $(0, 1)$, confirming strict convexity. 2. **Jensen's Bound**: \[ \frac{f(a) + f(b) + f(c)}{3} \ge f\left(\frac{a+b+c}{3}\right) = f\left(\frac{1}{3}\right) = \frac{1/3}...
Chosen establishes second-derivative convexity $f''(x) > 0$ and applies Jensen to hit the exact tight bound; Rejected uses a loose Cauchy-Schwarz.
How does Karamata's Majorization theorem generalize Jensen for sorted sequences?
<thought>Majorization x \succ y implies \sum f(x_i) \ge \sum f(y_i) for convex f.</thought> **Resolution**: Karamata proves that if $(x) \succ (y)$ in majorization order, $\sum f(x_i) \ge \sum f(y_i)$ for any convex function $f$.
math_r1_00048
Higher Algebra & Functional Equations
Higher Algebra
PUTNAM / RESEARCH
Competition Benchmark: Putnam Mathematical Competition [Problem #ALG-CHEBY-01047] Domain: Higher Algebra & Polynomials Problem Statement: Derive the closed-form expression for the Chebyshev polynomial of the first kind $T_{7}(x)$ satisfying $T_{7}(\cos \theta) = \cos(7\theta)$. Find all real roots of $T_{7}(x) = 0$ in ...
<thought> 1. Recurrence Relation: - $T_0(x) = 1, T_1(x) = x$. - $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$ from $\cos((n+1)\theta) + \cos((n-1)\theta) = 2\cos\theta\cos(n\theta)$. 2. Exact Roots: - $\cos(n\theta) = 0 \implies n\theta_k = \frac{2k-1}{2}\pi \implies \theta_k = \frac{2k-1}{2n}\pi$ for $k = 1, \dots, n...
<thought> 1. Recurrence Relation: - $T_0(x) = 1, T_1(x) = x$. - $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$ from $\cos((n+1)\theta) + \cos((n-1)\theta) = 2\cos\theta\cos(n\theta)$. 2. Exact Roots: - $\cos(n\theta) = 0 \implies n\theta_k = \frac{2k-1}{2}\pi \implies \theta_k = \frac{2k-1}{2n}\pi$ for $k = 1, \dots, n...
Since $T_n(x)$ is a polynomial of degree 7, its roots are equally spaced numbers between -1 and 1: $x_k = -1 + \frac{2k}{7}$, which ignores the non-linear cosine density.
**Derivation & Spectral Decomposition:** 1. **Recurrence**: $T_{n+1}(x) = 2x T_n(x) - T_{n-1}(x)$. 2. **Trigonometric Closed Form**: $T_{7}(x) = \cos(7 \arccos x)$ for $x \in [-1, 1]$. 3. **Roots**: Setting $\cos(7\theta) = 0$ gives $n$ distinct real roots: \[ \boxed{x_k = \cos\left(\frac{2k - 1}{2 \cdot 7} \pi\right)...
Chosen derives exact non-linear Chebyshev nodes $x_k = \cos(\frac{2k-1}{2n}\pi)$; Rejected assumes a naive uniform arithmetic progression.
Why do Chebyshev nodes minimize the Runge interpolation error phenomenon?
<thought>Equioscillation and minimization of the monic polynomial supremum norm ||P||_\infty = 2^{1-n}.</thought> **Resolution**: Chebyshev roots minimize the product $\prod |x - x_k|$, making the monic polynomial norm $\|T_n\|_\infty = 2^{1-n}$ minimal on $[-1, 1]$.
math_r1_00049
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-AMGM-01048] Domain: Formal Logic & Real Analysis Theorem Statement: Formalize and prove in Lean 4 that for non-negative reals $a, b$, $2\sqrt{ab} \le a + b$.
<thought> 1. Lean 4 Formalization Architecture: - Target Theorem: $\forall (a \, b : \mathbb{R}), 0 \le a \to 0 \le b \to 2 \cdot \sqrt{a \cdot b} \le a + b$. - Algebraic identity in Lean: $(a + b) - 2\sqrt{ab} = (\sqrt{a} - \sqrt{b})^2 \ge 0$. - Tactics: `have`, `calc`, `sq_nonneg`, `Real.sq_sqrt`, `linarith`...
<thought> 1. Lean 4 Formalization Architecture: - Target Theorem: $\forall (a \, b : \mathbb{R}), 0 \le a \to 0 \le b \to 2 \cdot \sqrt{a \cdot b} \le a + b$. - Algebraic identity in Lean: $(a + b) - 2\sqrt{ab} = (\sqrt{a} - \sqrt{b})^2 \ge 0$. - Tactics: `have`, `calc`, `sq_nonneg`, `Real.sq_sqrt`, `linarith`...
```lean4 theorem am_gm_two (a b : ℝ) (ha : 0 ≀ a) (hb : 0 ≀ b) : 2 * Real.sqrt (a * b) ≀ a + b := by have h := (a - b)^2 linarith ```
```lean4 import Mathlib.Data.Real.Basic import Mathlib.Data.Real.Sqrt import Mathlib.Tactic.Positivity import Mathlib.Tactic.Linarith /-- AM-GM Inequality for n = 2 in Lean 4 -/ theorem am_gm_two (a b : ℝ) (ha : 0 ≀ a) (hb : 0 ≀ b) : 2 * Real.sqrt (a * b) ≀ a + b := by have h_sq : 0 ≀ (Real.sqrt a - Real.sqrt b)...
Chosen provides a complete Mathlib4 calc chain with square root transformations; Rejected misuses $(a-b)^2$ which cannot eliminate square roots.
How do you prove strict equality condition $2\sqrt{ab} = a + b \iff a = b$?
<thought>By sq_eq_zero_iff and Real.sqrt_inj.</thought> **Resolution**: Equality holds when $(\sqrt{a} - \sqrt{b})^2 = 0 \iff \sqrt{a} = \sqrt{b} \iff a = b$.
math_r1_00050
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-BERNOULLI-01049] Domain: Formal Logic & Analysis Theorem Statement: Formalize and prove Bernoulli's Inequality in Lean 4 for all $n : \mathbb{N}$ and $x \ge -1$: \[ (1 + x)^n \ge 1 + n x \]
<thought> 1. Mathematical Induction in Lean 4: - Base case n = 0: (1+x)^0 = 1 \ge 1 + 0 = 1. Trivial by ring / norm_num. - Inductive step: Assume (1+x)^k \ge 1 + kx. - Multiply by (1+x) \ge 0: (1+x)^{k+1} \ge (1 + kx)(1 + x) = 1 + (k+1)x + k x^2 \ge 1 + (k+1)x. - Tactics: induction', linarith, nlinarith, mu...
<thought> 1. Mathematical Induction in Lean 4: - Base case n = 0: (1+x)^0 = 1 \ge 1 + 0 = 1. Trivial by ring / norm_num. - Inductive step: Assume (1+x)^k \ge 1 + kx. - Multiply by (1+x) \ge 0: (1+x)^{k+1} \ge (1 + kx)(1 + x) = 1 + (k+1)x + k x^2 \ge 1 + (k+1)x. - Tactics: induction', linarith, nlinarith, mu...
```lean4 theorem bernoulli_inequality (x : ℝ) (n : β„•) : 1 + (n : ℝ) * x ≀ (1 + x) ^ n := by induction' n with k hk Β· simp Β· nlinarith ```
```lean4 import Mathlib.Data.Real.Basic import Mathlib.Tactic.Linarith import Mathlib.Tactic.Ring /-- Bernoulli's Inequality by induction in Lean 4 -/ theorem bernoulli_inequality (x : ℝ) (hx : -1 ≀ x) (n : β„•) : 1 + (n : ℝ) * x ≀ (1 + x) ^ n := by induction' n with k hk Β· simp Β· have h_pos : 0 ≀ 1 + x := by ...
Chosen establishes $0 \le 1+x$ positivity and proves inductive step via calc; Rejected omits hypothesis $x \ge -1$ making the theorem false for negative odd powers.
Why is the hypothesis $x \ge -1$ essential in the Lean 4 proof?
<thought>Multiplying inequalities by a negative term reverses the inequality sign.</thought> **Resolution**: If $1+x < 0$, `mul_le_mul_of_nonneg_right` fails because multiplication by a negative value reverses $\le$ to $\ge$.
math_r1_00051
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-SQRT2-01050] Domain: Formal Logic & Number Theory Theorem Statement: Formalize and prove in Lean 4 that $\sqrt{2}$ is irrational by showing there exist no coprime natural numbers $p, q$ with $p^2 = 2q^2$.
<thought> 1. Classical Parity Proof in Lean 4: - Target: `theorem sqrt_two_irrational (p q : β„•) (hq : q β‰  0) (h_coprime : Nat.Coprime p q) : p ^ 2 β‰  2 * q ^ 2`. - If $p^2 = 2q^2$, then $2 \mid p^2 \implies 2 \mid p$. - Write $p = 2k$, then $4k^2 = 2q^2 \implies q^2 = 2k^2 \implies 2 \mid q$. - Contradiction...
<thought> 1. Classical Parity Proof in Lean 4: - Target: `theorem sqrt_two_irrational (p q : β„•) (hq : q β‰  0) (h_coprime : Nat.Coprime p q) : p ^ 2 β‰  2 * q ^ 2`. - If $p^2 = 2q^2$, then $2 \mid p^2 \implies 2 \mid p$. - Write $p = 2k$, then $4k^2 = 2q^2 \implies q^2 = 2k^2 \implies 2 \mid q$. - Contradiction...
```lean4 theorem sqrt_two_irrational (p q : β„•) : p ^ 2 β‰  2 * q ^ 2 := by intro h have h2 : p = 0 := by linarith sorry ```
```lean4 import Mathlib.Data.Nat.Prime import Mathlib.Tactic.Ring /-- Proof of irrationality of sqrt(2) in Lean 4 -/ theorem sqrt_two_irrational (p q : β„•) (h_coprime : p.Coprime q) : p ^ 2 β‰  2 * q ^ 2 := by intro h_eq have h_even_p2 : 2 ∣ p ^ 2 := by use q ^ 2 exact h_eq.symm have h_even_p : 2 ∣ p :=...
Chosen uses formal Mathlib4 `Nat.Prime.dvd_of_dvd_pow` to prove parity contradiction on coprime gcd; Rejected collapses to sorry.
How does this generalize to $\sqrt{n}$ for any non-square integer $n$?
<thought>By p-adic valuation v_p(p^2) = 2 v_p(p) is always even, while v_p(n q^2) is odd for prime p with odd exponent in n.</thought> **Resolution**: Fundamental Theorem of Arithmetic implies $v_p(p^2)$ is even, whereas $v_p(n q^2) = v_p(n) + 2v_p(q)$ is odd, yielding an immediate contradiction.
math_r1_00052
Formal Theorem Proving in Lean 4
Formal Theorem Proving in Lean 4
LEAN 4 FORMAL PROOF
Benchmark: Lean 4 Formal Verification (Mathlib4) [Task #LEAN4-LAGRANGE-01051] Domain: Formal Logic & Abstract Algebra Theorem Statement: Formalize in Lean 4 the coset partition equality: for any finite group $G$ and subgroup $H \le G$, $|G| = [G : H] \cdot |H|$.
<thought> 1. Lean 4 Abstract Algebra Architecture: - Typeclasses: `Group G`, `Fintype G`, `Subgroup H`. - Left coset equivalence relation $a \sim b \iff a^{-1}b \in H$. - Bijections between distinct cosets: $x H \cong H$ via $h \mapsto x h$. </thought>
<thought> 1. Lean 4 Abstract Algebra Architecture: - Typeclasses: `Group G`, `Fintype G`, `Subgroup H`. - Left coset equivalence relation $a \sim b \iff a^{-1}b \in H$. - Bijections between distinct cosets: $x H \cong H$ via $h \mapsto x h$. </thought> ```lean4 import Mathlib.Algebra.Group.Subgroup.Basic impo...
```lean4 theorem lagrange_group_order {G : Type*} [Group G] (H : Subgroup G) : Fintype.card G = Fintype.card H := by rfl ```
```lean4 import Mathlib.Algebra.Group.Subgroup.Basic import Mathlib.GroupTheory.Index import Mathlib.Data.Fintype.Card /-- Lagrange's Theorem for Finite Groups in Lean 4 -/ theorem lagrange_group_order {G : Type*} [Group G] [Fintype G] (H : Subgroup G) [Fintype H] : Fintype.card G = H.index * Fintype.card H := by ...
Chosen applies correct Mathlib4 subgroup quotient cardinality theorem; Rejected falsely claims subgroup order equals group order by rfl.
What is the corollary for prime-order groups?
<thought>If |G| = p, proper subgroups have order 1 or p, forcing G to be cyclic.</thought> **Resolution**: Every group of prime order $p$ has no non-trivial proper subgroups, and is therefore cyclic ($G \cong \mathbb{Z}/p\mathbb{Z}$).
math_r1_00053
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
PUTNAM / RESEARCH
Competition Benchmark: USAMO National Olympiad [Problem #COMB-RAMSEY-01052] Domain: Advanced Combinatorics Problem Statement: In complete graph $K_n$, prove using the probabilistic method with ErdΕ‘s expectation and the Symmetric LovΓ‘sz Local Lemma (LLL) that diagonal Ramsey number satisfies $R(5, 5) > \frac{\sqrt{2}5}{...
<thought> 1. Random Edge Coloring: - Each edge colored Red/Blue with $p = 1/2$. - Probability of monochromatic $K_k$: $P(A_S) = 2^{1 - \binom{k}{2}}$. 2. Dependency Graph & LLL: - Two subsets $S, T$ are dependent iff $|S \cap T| \ge 2$. - Degree $d = \binom{k}{2} \binom{n-2}{k-2}$. - LLL Condition: $e \c...
<thought> 1. Random Edge Coloring: - Each edge colored Red/Blue with $p = 1/2$. - Probability of monochromatic $K_k$: $P(A_S) = 2^{1 - \binom{k}{2}}$. 2. Dependency Graph & LLL: - Two subsets $S, T$ are dependent iff $|S \cap T| \ge 2$. - Degree $d = \binom{k}{2} \binom{n-2}{k-2}$. - LLL Condition: $e \c...
Color alternating edges Red and Blue. In any graph, there are at most $n/2$ edges of one color, so no monochromatic clique of size 5 can exist for $n < 2^5$.
**Probabilistic Bound via Symmetric LovΓ‘sz Local Lemma:** 1. **Bad Event Probability**: For subset $S$ of size $5$, $P(A_S) = 2^{1 - \binom{5}{2}}$. 2. **Dependency Degree**: $d = \binom{5}{2} \binom{n-2}{5-2} < \frac{5^2}{2} \frac{n^{5-2}}{(5-2)!}$. 3. **LLL Criterion**: $e \cdot P(A_S) \cdot (d + 1) \le 1$ ensures $...
Chosen derives the exact dependency degree $d$ and invokes Symmetric LLL; Rejected makes an invalid deterministic alternating coloring claim.
How does Spencer's polynomial method compare with LLL?
<thought>Spencer's fractional coloring achieves optimal constants within the probabilistic method.</thought> **Resolution**: Spencer's theorem matches the asymptotic constant while providing an algorithmic Moser-Tardos resample algorithm.
math_r1_00054
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
AIME / USAMO
Competition Benchmark: Putnam Mathematical Competition [Problem #COMB-DERANGE-01053] Domain: Advanced Combinatorics Problem Statement: Using the Principle of Inclusion-Exclusion, derive the exact closed formula for derangements $D_n = !8$ (permutations of $8$ elements with no fixed points) and evaluate $D_{8}$ exactly.
<thought> 1. Inclusion-Exclusion Formulation: - Let $A_i$ be permutations fixing $i$. $|A_i| = (n-1)!$. - Intersection of $k$ sets: $|igcap_{j=1}^k A_{i_j}| = (n-k)!$. - $D_n = n! - \sum |A_i| + \sum |A_i \cap A_j| - \dots = n! \sum_{k=0}^n \frac{(-1)^k}{k!}$. 2. Evaluation for n = 8: - $D_{n} = \lfloor \f...
<thought> 1. Inclusion-Exclusion Formulation: - Let $A_i$ be permutations fixing $i$. $|A_i| = (n-1)!$. - Intersection of $k$ sets: $|igcap_{j=1}^k A_{i_j}| = (n-k)!$. - $D_n = n! - \sum |A_i| + \sum |A_i \cap A_j| - \dots = n! \sum_{k=0}^n \frac{(-1)^k}{k!}$. 2. Evaluation for n = 8: - $D_{n} = \lfloor \f...
Since each element has $n-1$ choices for its position, $D_n = (n-1)^n$. For $n=8$, this gives $(8-1)^{8}$, which ignores dependency among cycles.
**Derivation via Inclusion-Exclusion:** 1. **Union of Fixed Points**: $|\bigcup_{i=1}^{8} A_i| = \sum_{k=1}^{8} (-1)^{k-1} \binom{8}{k} (8-k)! = 8! \sum_{k=1}^{8} \frac{(-1)^{k-1}}{k!}$. 2. **Complement**: $D_{8} = 8! \left(1 - \sum_{k=1}^{8} \frac{(-1)^{k-1}}{k!}\right) = 8! \sum_{k=0}^{8} \frac{(-1)^k}{k!}$. \boxed...
Chosen uses formal PIE with alternating sum converging to $n!/e$; Rejected uses independent choices $(n-1)^n$ allowing duplicate assignments.
How does the recurrence $D_n = (n-1)(D_{n-1} + D_{n-2})$ relate to permutation cycle decomposition?
<thought>Element 1 swaps with element k (D_{n-2}) or is part of a larger cycle (D_{n-1}).</thought> **Resolution**: Partitioning based on whether element 1 belongs to a 2-cycle or cycle of length $\ge 3$ yields $D_n = (n-1)(D_{n-1} + D_{n-2})$.
math_r1_00055
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
PUTNAM / IMO HARD
Competition Benchmark: Lean 4 Mathlib Research [Problem #COMB-CAYLEY-01054] Domain: Advanced Combinatorics Problem Statement: Prove Cayley's Formula for the number of labeled trees on $11$ vertices $T_{11} = 11^{11-2}$ by establishing a bijective correspondence with PrΓΌfer sequences of length $11-2$.
<thought> 1. PrΓΌfer Sequence Bijection: - Tree \to PrΓΌfer: Successively remove leaf with smallest label and record its neighbor. Repeat until 2 vertices remain. - Result is a sequence of length n - 2 with entries in \{1, \dots, n\}. - Number of such sequences is n^{n-2}. - Inverse algorithm reconstructs uni...
<thought> 1. PrΓΌfer Sequence Bijection: - Tree \to PrΓΌfer: Successively remove leaf with smallest label and record its neighbor. Repeat until 2 vertices remain. - Result is a sequence of length n - 2 with entries in \{1, \dots, n\}. - Number of such sequences is n^{n-2}. - Inverse algorithm reconstructs uni...
A tree has $n-1$ edges chosen from $\binom{n}{2}$ possible edges. So $T_n = \binom{\binom{n}{2}}{n-1}$, which counts all acyclic and cyclic subgraphs indiscriminately.
**Bijective Proof via PrΓΌfer Sequences:** 1. **Tree to Sequence**: Iteratively delete the minimal leaf $v$ and append its unique neighbor $u$ to sequence $P$. The process terminates at length $11-2$. 2. **Reconstruction**: Each vertex appears in $P$ exactly $\text{deg}(v) - 1$ times. The minimal missing label is the f...
Chosen establishes explicit bijection with PrΓΌfer sequences; Rejected naively counts arbitrary subsets of edges $\binom{N}{n-1}$.
How does Matrix-Tree Theorem prove Cayley's formula algebraically?
<thought>Laplacian matrix of K_n has eigenvalues n with multiplicity n-1, so det cofactor is n^{n-2}.</thought> **Resolution**: The Laplacian of $K_n$ has spectrum $\{0, n, \dots, n\}$, and any cofactor equals $\frac{1}{n} \prod_{i=1}^{n-1} n = n^{n-2}$.
math_r1_00056
Advanced Combinatorics & Graph Invariants
Advanced Combinatorics
AIME / USAMO
Competition Benchmark: Putnam Mathematical Competition [Problem #COMB-CATALAN-01055] Domain: Advanced Combinatorics Problem Statement: Using Ordinary Generating Functions (OGF), solve the Catalan recurrence $C_0 = 1, C_{n+1} = \sum_{i=0}^n C_i C_{n-i}$ and compute $C_{9}$.
<thought> 1. OGF Setup: - Let C(x) = \sum_{n=0}^\infty C_n x^n. - C(x) = 1 + x C(x)^2 \implies x C(x)^2 - C(x) + 1 = 0. - Roots: C(x) = \frac{1 \pm \sqrt{1 - 4x}}{2x}. - Since C(0) = 1, we select minus sign: C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}. 2. Taylor Expansion: - \sqrt{1-4x} = 1 - \sum_{n=1}^\infty \...
<thought> 1. OGF Setup: - Let C(x) = \sum_{n=0}^\infty C_n x^n. - C(x) = 1 + x C(x)^2 \implies x C(x)^2 - C(x) + 1 = 0. - Roots: C(x) = \frac{1 \pm \sqrt{1 - 4x}}{2x}. - Since C(0) = 1, we select minus sign: C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}. 2. Taylor Expansion: - \sqrt{1-4x} = 1 - \sum_{n=1}^\infty \...
The recurrence $C_{n+1} = \sum C_i C_{n-i}$ is solved by $C_n = 2^n$ because at each step we make a binary choice.
**Closed Form via Algebraic Generating Function:** 1. **Functional Equation**: $C(x) = 1 + x C(x)^2 \implies C(x) = \frac{1 - \sqrt{1 - 4x}}{2x}$. 2. **Binomial Series**: $\sqrt{1 - 4x} = 1 - 2 \sum_{n=1}^\infty \frac{1}{n} \binom{2n-2}{n-1} x^n$. 3. **Extraction**: $C_n = \frac{1}{n+1} \binom{2n}{n}$. \boxed{C_{9} =...
Chosen derives algebraic OGF $C(x) = (1-\sqrt{1-4x})/2x$ and extracts exact coefficients; Rejected mistakes Catalan growth for exponential $2^n$.
How does Dyck path reflection principle prove $C_n = \binom{2n}{n} - \binom{2n}{n+1}$?
<thought>Reflecting bad paths that touch y = -1 across the line y = -1 establishes a bijection to paths ending at (2n, -2).</thought> **Resolution**: AndrΓ©'s Reflection Principle reflects the portion of an invalid path after its first touch of $y=-1$, mapping invalid paths bijectively to $\binom{2n}{n-1}$.
math_r1_00057
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: AIME 2026 Invitational [Problem #GEOM-EULER-01056] Domain: Euclidean Geometry Problem Statement: In triangle $\triangle ABC$, inradius is $r = 30$ and circumradius is $R = 96$. Prove Euler's formula $d^2 = R(R - 2r)$ for distance $d = IO$ and compute $d$.
<thought> 1. Geometric Invariant: - Power of incenter $I$ with respect to circumcircle: $R^2 - d^2 = AI \cdot IM$. - Trillium Theorem: $IM = MB$. - $AI = r / \sin(A/2)$, $MB = 2R \sin(A/2)$. - $AI \cdot MB = 2Rr \implies d^2 = R(R - 2r) = 3456$. </thought>
<thought> 1. Geometric Invariant: - Power of incenter $I$ with respect to circumcircle: $R^2 - d^2 = AI \cdot IM$. - Trillium Theorem: $IM = MB$. - $AI = r / \sin(A/2)$, $MB = 2R \sin(A/2)$. - $AI \cdot MB = 2Rr \implies d^2 = R(R - 2r) = 3456$. </thought> **Synthetic Proof via Trillium Lemma:** 1. **Powe...
By Pythagoras theorem in $\triangle AIO$, $d^2 = R^2 + r^2$. For $R=96, r=30$, $d = \sqrt{10116}$, which ignores non-orthogonality.
**Synthetic Proof via Trillium Lemma:** 1. **Power of Incenter**: $R^2 - d^2 = AI \cdot IM$. 2. **Trillium Lemma**: $IM = MB = 2R \sin(A/2)$ and $AI = \frac{r}{\sin(A/2)}$. 3. **Product**: $AI \cdot IM = 2Rr \implies \boxed{d^2 = R(R - 2r) = 3456 \implies d = \sqrt{3456}}$.
Chosen applies Trillium Lemma and Power of Point on circumcircle; Rejected falsely assumes $OI \perp AI$ applying Pythagorean theorem.
How does $d^2 \ge 0$ immediately prove Euler's inequality $R \ge 2r$?
<thought>Real Euclidean distance requires d^2 >= 0, so R(R-2r) >= 0.</thought> **Resolution**: Since $R > 0$, $R(R-2r) = d^2 \ge 0$ yields $R \ge 2r$, with equality $R = 2r$ if and only if $d = 0$ (equilateral triangle).
math_r1_00058
Euclidean & Projective Geometry
Euclidean
USAMO / IMO
Competition Benchmark: Putnam Mathematical Competition [Problem #GEOM-PTOLEMY-01057] Domain: Euclidean & Projective Geometry Problem Statement: In cyclic quadrilateral $ABCD$ with sides $AB=7, BC=8, CD=10, DA=4$, prove Ptolemy's Theorem $AC \cdot BD = AB \cdot CD + BC \cdot DA$ using circular inversion.
<thought> 1. Geometric Inversion Method: - Invert with center A and arbitrary radius R. - Circle passing through A, B, C, D maps to a straight line containing B', C', D' in order. - Segment lengths under inversion: X'Y' = \frac{R^2 \cdot XY}{AX \cdot AY}. - Collinear segments: B'D' = B'C' + C'D'. - Subst...
<thought> 1. Geometric Inversion Method: - Invert with center A and arbitrary radius R. - Circle passing through A, B, C, D maps to a straight line containing B', C', D' in order. - Segment lengths under inversion: X'Y' = \frac{R^2 \cdot XY}{AX \cdot AY}. - Collinear segments: B'D' = B'C' + C'D'. - Subst...
Apply law of cosines to $\triangle ABC$ and $\triangle ADC$. Then add the two diagonal formulas directly without matching opposite supplementary angles.
**Proof via Circular Inversion:** 1. **Inversion Action**: Invert the figure at center $A$. The circumcircle through $A$ maps to a straight line on which points $B', C', D'$ lie in collinear order. 2. **Metric Transformation**: $B'D' = \frac{R^2 \cdot BD}{AB \cdot AD}$, $B'C' = \frac{R^2 \cdot BC}{AB \cdot AC}$, $C'D'...
Chosen establishes rigorous proof via circular inversion mapping circles to lines; Rejected fails to eliminate cosines of opposite angles.
How does Ptolemy's inequality hold for non-cyclic quadrilaterals?
<thought>Triangle inequality on the inverted points B', C', D' gives B'D' <= B'C' + C'D'.</thought> **Resolution**: For non-cyclic quadrilaterals, $B', C', D'$ form a non-degenerate triangle, so triangle inequality yields strict $AC \cdot BD < AB \cdot CD + BC \cdot DA$.
math_r1_00059
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: Lean 4 Mathlib Research [Problem #GEOM-RADICAL-01058] Domain: Euclidean Geometry Problem Statement: Let $\omega_1, \omega_2, \omega_3$ be three non-concentric circles whose centers are not collinear. Prove that the three radical axes pairwise intersect at a unique point $P$ (the Radical Center).
<thought> 1. Power of a Point & Radical Axis Definition: - For circle \omega(O, R), \text{Pow}(X, \omega) = d(X, O)^2 - R^2. - Radical axis L_{12} = \{X : \text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)\}. 2. Intersection Proof: - Let P = L_{12} \cap L_{23}. - Then \text{Pow}(P, \omega_1) = \text{Pow}(P, ...
<thought> 1. Power of a Point & Radical Axis Definition: - For circle \omega(O, R), \text{Pow}(X, \omega) = d(X, O)^2 - R^2. - Radical axis L_{12} = \{X : \text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)\}. 2. Intersection Proof: - Let P = L_{12} \cap L_{23}. - Then \text{Pow}(P, \omega_1) = \text{Pow}(P, ...
Since all circles are round, their chords intersect at the center of mass of the three centers.
**Proof of Radical Center Concurrence:** 1. **Linearity of Radical Axis**: The locus $\text{Pow}(X, \omega_1) = \text{Pow}(X, \omega_2)$ is a straight line perpendicular to the line of centers $O_1 O_2$. 2. **Intersection**: Since centers are non-collinear, line $L_{12}$ and $L_{23}$ intersect at unique point $P$. 3. ...
Chosen applies analytic Power of Point transitivity $\text{Pow}(P, \omega_1) = \text{Pow}(P, \omega_3)$; Rejected makes a nonsensical center-of-mass heuristic.
What happens if the centers $O_1, O_2, O_3$ are collinear?
<thought>The radical axes are parallel lines perpendicular to the line of centers.</thought> **Resolution**: If centers are collinear, the three radical axes are mutually parallel perpendiculars (intersecting at infinity in projective plane).
math_r1_00060
Euclidean & Projective Geometry
Euclidean
IMO SHORTLIST
Competition Benchmark: IMO Shortlist Benchmark [Problem #GEOM-SIMSON-01059] Domain: Euclidean & Projective Geometry Problem Statement: Prove the Simson Line Theorem: The orthogonal projections $P_a, P_b, P_c$ of point $P$ onto sides $BC, CA, AB$ of $\triangle ABC$ are collinear if and only if $P$ lies on the circumcirc...
<thought> 1. Cyclic Quadrilateral Angle Chasing: - Points P, P_a, B, P_c are concyclic on circle with diameter PB. - Points P, P_a, C, P_b are concyclic on circle with diameter PC. - Collinearity of P_a, P_b, P_c requires \angle P P_a P_c = \angle P P_a P_b. - In cyclic quads: \angle P P_a P_c = \angle P B ...
<thought> 1. Cyclic Quadrilateral Angle Chasing: - Points P, P_a, B, P_c are concyclic on circle with diameter PB. - Points P, P_a, C, P_b are concyclic on circle with diameter PC. - Collinearity of P_a, P_b, P_c requires \angle P P_a P_c = \angle P P_a P_b. - In cyclic quads: \angle P P_a P_c = \angle P B ...
Connect the three pedal points to form $\triangle P_a P_b P_c$. Use Heron's formula for area. Setting area to 0 gives $P_a = P_b$, which requires $P$ to be at a vertex.
**Angle-Chasing Proof of Simson's Theorem:** 1. **Concyclic Pedal Quadrilaterals**: Since $\angle P P_a B = \angle P P_c B = 90^\circ$, quadrilateral $P P_a B P_c$ is cyclic with diameter $PB$. Likewise, $P P_a C P_b$ is cyclic with diameter $PC$. 2. **Inscribed Angle Equivalence**: $\angle P P_a P_c = \angle P B A$ a...
Chosen uses cyclic pedal circles to prove $\angle PBA = \angle PCA$ iff $P$ on circumcircle; Rejected confuses collinearity with point collapse $P_a = P_b$.
How does the Simson line bisect the segment $P H$ where $H$ is the orthocenter?
<thought>Homothety of ratio 1/2 from H maps circumcircle to nine-point circle, placing midpoint of PH on Simson line.</thought> **Resolution**: The midpoint of $PH$ lies on the nine-point circle and the Simson line of $P$, meaning the Simson line bisects $PH$.
math_r1_00061
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
AIME / IMO HARD
Competition Benchmark: Putnam Mathematical Competition [Problem #NT-PELL-01060] Domain: Olympiad Number Theory Problem Statement: Find all positive integer pairs $(x, y)$ such that $x^2 - 11y^2 = 1$ with $x \equiv 111 \pmod{149}$. Prove that there are infinitely many such pairs and determine the periodicity of solution...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{11}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 149 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(11\tilde{y}_1^2)$ is a quadratic non...
<thought> 1. Initial Pell Structure: - Fundamental solution $(\tilde{x}_1, \tilde{y}_1) \in \mathbb{Z}[\sqrt{11}]^\times$. - Recurrence $x_{n+1} = 2\tilde{x}_1 x_n - x_{n-1}$. 2. Modulo 149 Reciprocity & Periodicity Backtracking: - Wait, what if the discriminant $\Delta = 4(11\tilde{y}_1^2)$ is a quadratic non...
Substitute $x=111$ directly into $x^2 - 11y^2 = 1$. Then $y = \sqrt{(111^2-1)/11}$. Since $111^2-1$ can be made divisible by $11$, solutions exist without checking group periodicity.
**Theorem & Complete Algebraic Number Theory Solution:** 1. **Pell Group Action**: Solutions satisfy $x_n + y_n\sqrt{11} = (x_1 + y_1\sqrt{11})^n$. 2. **Periodicity**: Over $\mathbb{F}_{149}$, if $(11/149) = 1$, period divides $p-1$; if $(11/149) = -1$, period divides $p+1$. 3. **Infinitude**: The residue $x_{n_0} \eq...
Chosen proves periodicity in algebraic number field and analyzes quadratic character; Rejected naively evaluates a single value without proving infinitude or periodicity.
How does the norm map $N(\alpha) = 1$ in $\mathbb{F}_{151}$ force the period to divide $p+1$?
<thought>Norm map kernel has order $(p^2-1)/(p-1) = p+1$.</thought> **Lemma**: The cyclic kernel of $N: \mathbb{F}_{151}^\times \to \mathbb{F}_{149}^\times$ has exact cardinality $p+1$.
math_r1_00062
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
IMO SHORTLIST
Competition Benchmark: IMO Shortlist Benchmark [Problem #NT-LTE-01061] Domain: Olympiad Number Theory Problem Statement: Let $p = 3$ be an odd prime. Suppose positive integers $a, b$ satisfy $p \mid (a - b)$ but $p \nmid a$ and $p \nmid b$. Prove the Lifting The Exponent (LTE) identity for $n = 15$: \[ v_{3}(a^{15} - b...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
<thought> 1. LTE Lemma Exploration: - Factorization: $a^n - b^n = (a - b) \sum_{k=0}^{n-1} a^{n-1-k} b^k$. - Let $S = \sum_{k=0}^{n-1} a^{n-1-k} b^k$. We must prove $v_p(S) = v_p(n)$. 2. Backtracking & Binomial Expansion: - Write $a = b + p^k m$ where $k = v_p(a-b) \ge 1$. - $a^j = (b + p^k m)^j \equiv b^j ...
By expanding $a^n - b^n = (a-b)(a^{n-1} + \dots + b^{n-1})$, each of the $n$ terms is equal to $a^{n-1} \pmod p$, so the sum is $n a^{n-1}$. Thus $v_p(a^n - b^n) = v_p(a-b) + v_p(n a^{n-1}) = v_p(a-b) + v_p(n) + (n-1)v_p(a)$, which gives an extra term.
**Proof via Binomial p-adic Expansion:** 1. Let $v_p(a - b) = k \ge 1$. Write $a = b + c$ with $v_p(c) = k$. 2. Expanding $a^n = (b + c)^n = b^n + n b^{n-1} c + \sum_{j=2}^n \binom{n}{j} b^{n-j} c^j$. 3. For $j \ge 2$, $v_p(\binom{n}{j} c^j) \ge j k - v_p(j) > k + v_p(n)$ for odd prime $p$. 4. Therefore $v_p(a^n - b^n...
Chosen establishes p-adic valuations with strict inequality $j k - v_p(j) > k + v_p(n)$; Rejected incorrectly asserts $v_p(a) > 0$ when $p \nmid a$.
Why does LTE fail for $p = 2$ when $n$ is even?
<thought>For p=2 and j=2, 2k - 1 = k + (k-1) equals k when k=1, causing cancellation.</thought> **Resolution**: For $p=2$, the second-order binomial coefficient has $v_2(\binom{n}{2} c^2) = v_2(n) - 1 + 2k$, requiring the extra correction term $v_2(a+b)-1$.
math_r1_00063
Olympiad Number Theory & Diophantine Equations
Olympiad Number Theory
PUTNAM / IMO HARD
Competition Benchmark: AIME 2026 Invitational [Problem #NT-GAUSS-01062] Domain: Olympiad Number Theory Problem Statement: Evaluate the Legendre symbol $\left(\frac{73}{43}\right)$ using Gauss's Law of Quadratic Reciprocity. Prove whether the congruence $x^2 \equiv 73 \pmod{43}$ is solvable in integers.
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 73 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $43 \pmod{73} = 43$. - Decompose into prime factors and apply Euler's criterion. </thou...
<thought> 1. Quadratic Reciprocity Law: - $(p/q)(q/p) = (-1)^{(p-1)/2 \cdot (q-1)/2}$. - Here $p = 73 \equiv 1 \pmod 4$, so $(p-1)/2$ is even. - Therefore $(p/q) = (q/p)$. 2. Reduction Modulo p: - Compute $q \pmod p$: $43 \pmod{73} = 43$. - Decompose into prime factors and apply Euler's criterion. </thou...
Since both 73 and 43 are odd primes, $(p/q) = -(q/p)$ always holds by Gauss's theorem. Evaluating directly gives the opposite sign.
**Evaluation via Quadratic Reciprocity:** 1. **Parity Check**: Since $p = 73 \equiv 1 \pmod 4$, $(-1)^{\frac{73-1}{2}\frac{43-1}{2}} = 1$. 2. **Reciprocity Application**: $\left(\frac{73}{43}\right) = \left(\frac{43}{73}\right)$. 3. **Modular Reduction**: Evaluating $\left(\frac{43}{73}\right) = -1$. Therefore, the c...
Chosen correctly analyzes $p \equiv 1 \pmod 4$ to obtain $(p/q) = +(q/p)$; Rejected misremembers Quadratic Reciprocity as always having a negative sign.
How does Gauss's Lemma with half-intervals prove quadratic reciprocity?
<thought>Counting lattice points in rectangle of size (p-1)/2 by (q-1)/2.</thought> **Resolution**: Eisenstein's geometric proof counts interior integer points in the rectangle $[1, (p-1)/2] \times [1, (q-1)/2]$.
End of preview. Expand in Data Studio

πŸ“ Enterprise DeepSeek-R1 Autonomous Mathematical & Logic CoT SFT/DPO Dataset (2026)

Open In Colab License: CC BY 4.0 Domains Gumroad Suite

High-precision multi-turn instruction tuning and preference optimization dataset with step-by-step hypothesis exploration, error discovery, and dynamic backtracking Chain-of-Thought (<thought>) reasoning trees for fine-tuning LLMs (DeepSeek-R1-Distill-Qwen, Qwen-2.5-Math, Llama-3.3, Mistral) into World-Class Olympiad Mathematicians and Formal Verification Agents.


πŸ“Š Dataset Architecture & 20 Mathematical Cores

  • 20 Independent Theorem Cores: Olympiad Number Theory (Pell's Equations, LTE Lemma, Quadratic Reciprocity, Chinese Remainder & Hensel), Higher Algebra (Cauchy functional equations, Newton-Girard sums, Jensen convexity, Chebyshev recurrence), Formal Theorem Proving in Lean 4 (AM-GM, Bernoulli induction, Sqrt(2) irrationality, Lagrange group cosets), Advanced Combinatorics (Ramsey & LLL, Derangements & PIE, Cayley tree PrΓΌfer bijection, Catalan OGF/EGF), and Euclidean/Projective Geometry (Euler distance & Trillium, Ptolemy inversion, Radical axis coaxal systems, Simson line).
  • Dynamic DeepSeek-R1 Backtracking: Explores hypotheses -> detects contradictions ("Wait, let me rethink...") -> derives rigorous proof.
  • Hardware & VRAM Profiling: Pre-computed token quantiles ($P_{50} = 375$, $P_{99} = 497$) and GPU VRAM memory footprint matrix across model sizes.
  • DPO Preference Pairs: 2,500 curated Chosen vs Subtle Mathematical Traps with automated formal logic critiques.
  • Pre-packaged Tooling: 1-Click Unsloth Colab Notebook (1_CLICK_UNSLOTH_MATH_R1_FINE_TUNING_COLAB.ipynb), LLaMA-Factory YAML, and Axolotl YML configs.

πŸš€ Quick Start with Hugging Face datasets

from datasets import load_dataset

# Load from Parquet
dataset = load_dataset("parquet", data_files="DEEPSEEK_R1_MATH_LOGIC_COT_SFT_DPO_2026_100_SAMPLE.parquet", split="train")

print(f"Loaded {len(dataset)} sample mathematical training instances!")
print("Sample Problem:", dataset[0]["user_prompt"])
print("Sample Proof:", dataset[0]["formal_proof"])

πŸ‘‘ Get the Complete 10,000 SFT + 2,500 DPO Enterprise Suite

Need the complete production-grade package with 10,000 multi-turn SFT pairs across 20 theorem cores, 2,500 DPO pairs, pre-packaged Train/Val/Test splits, full 39.1 MB SQLite DB, and commercial enterprise deployment rights?

πŸ‘‰ Get Complete 10,000 SFT/DPO Suite on Gumroad

Downloads last month
130