id
string
sources
list
title
string
abstract
string
authors
list
categories
list
fields_of_study
list
published_date
timestamp[s]
url
string
pdf_url
string
arxiv_id
string
doi
string
citation_count
int64
influential_citation_count
int64
has_code
bool
code_url
string
venue
string
quality_score
float64
ff9c6512108e2752ffa33d59e2ad24b929d8ff640200e7080ee6d7fc26ae1435
[ "arxiv", "semantic_scholar" ]
Learning to Discover at Test Time
How can we use AI to discover a new state of the art for a scientific problem? Prior work in test-time scaling, such as AlphaEvolve, performs search by prompting a frozen LLM. We perform reinforcement learning at test time, so the LLM can continue to train, but now with experience specific to the test problem. This for...
[ "Mert Yuksekgonul", "Daniel Koceja", "Xinhao Li", "Federico Bianchi", "Jed McCaleb", "Xiaolong Wang", "Jan Kautz", "Yejin Choi", "James Zou", "Carlos Guestrin", "Yu Sun" ]
[ "cs.LG", "cs.AI" ]
[ "Computer Science" ]
2026-01-22T00:00:00
https://arxiv.org/abs/2601.16175
https://arxiv.org/pdf/2601.16175v2
2601.16175
10.48550/arXiv.2601.16175
56
7
true
https://github.com/test-time-training/discover
arXiv.org
0.6942
cb615dc6aedc12f9d4849e17fa9416e3a1ee6972bd8447b1988574c81fe1acd0
[ "arxiv", "semantic_scholar" ]
Render-of-Thought: Rendering Textual Chain-of-Thought as Images for Visual Latent Reasoning
Chain-of-Thought (CoT) prompting has achieved remarkable success in unlocking the reasoning capabilities of Large Language Models (LLMs). Although CoT prompting enhances reasoning, its verbosity imposes substantial computational overhead. Recent works often focus exclusively on outcome alignment and lack supervision on...
[ "Yifan Wang", "Shiyu Li", "Peiming Li", "Xiaochen Yang", "Yang Tang", "Zheng Wei" ]
[ "cs.CL", "cs.CV" ]
[ "Computer Science" ]
2026-01-21T00:00:00
https://arxiv.org/abs/2601.14750
https://arxiv.org/pdf/2601.14750v4
2601.14750
10.48550/arXiv.2601.14750
11
0
true
https://github.com/TencentBAC/RoT
arXiv.org
0.6924
5739aa73946037775f108d6745cdad30d604f97a8a55ff227f137728adcce536
[ "arxiv", "semantic_scholar" ]
The Geometry of Thought: How Scale Restructures Reasoning In Large Language Models
Scale does not uniformly improve reasoning - it restructures it. Analyzing 25,000+ chain-of-thought trajectories across four domains (Law, Science, Code, Math) and two scales (8B, 70B parameters), we discover that neural scaling laws trigger domain-specific phase transitions rather than uniform capability gains. Legal ...
[ "Samuel Cyrenius Anderson" ]
[ "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2026-01-19T00:00:00
https://arxiv.org/abs/2601.13358
https://arxiv.org/pdf/2601.13358v2
2601.13358
10.48550/arXiv.2601.13358
0
0
false
null
arXiv.org
0.4457
ce48e1a738779937c38c377103785a6660e2c9ee8873ff2c3d0d15580f3b2bf7
[ "arxiv", "semantic_scholar" ]
Agentic Reasoning for Large Language Models
Reasoning is a fundamental cognitive process underlying inference, problem-solving, and decision-making. While large language models (LLMs) demonstrate strong reasoning capabilities in closed-world settings, they struggle in open-ended and dynamic environments. Agentic reasoning marks a paradigm shift by reframing LLMs...
[ "Tianxin Wei", "Ting-Wei Li", "Zhining Liu", "Xuying Ning", "Ze Yang", "Jiaru Zou", "Zhichen Zeng", "Ruizhong Qiu", "Xiao Lin", "Dongqi Fu", "Zihao Li", "Mengting Ai", "Duo Zhou", "Wenxuan Bao", "Yunzhe Li", "Gaotang Li", "Cheng Qian", "Yu Wang", "Xiangru Tang", "Yin Xiao", "...
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2026-01-18T00:00:00
https://arxiv.org/abs/2601.12538
https://arxiv.org/pdf/2601.12538v1
2601.12538
10.48550/arXiv.2601.12538
27
4
true
https://github.com/weitianxin/Awesome-Agentic-Reasoning
arXiv.org
0.6871
752551990802e8f84e2023cd897024bc110379d08c30c55f7e591c919015c4c6
[ "arxiv", "semantic_scholar" ]
Neural Chain-of-Thought Search: Searching the Optimal Reasoning Path to Enhance Large Language Models
Chain-of-Thought reasoning has significantly enhanced the problem-solving capabilities of Large Language Models. Unfortunately, current models generate reasoning steps sequentially without foresight, often becoming trapped in suboptimal reasoning paths with redundant steps. In contrast, we introduce Neural Chain-of-Tho...
[ "Guoming Ling", "Zhongzhan Huang", "Yupei Lin", "Junxin Li", "Shanshan Zhong", "Hefeng Wu", "Liang Lin" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-01-16T00:00:00
https://arxiv.org/abs/2601.11340
https://arxiv.org/pdf/2601.11340v2
2601.11340
10.48550/arXiv.2601.11340
7
0
true
https://github.com/MilkThink-Lab/Neural-CoT-Search
arXiv.org
0.6835
8aeefb9ece36c06e74452aae31ae299734feca4fdd0bad0ea4c8b9933b0d5e00
[ "arxiv", "semantic_scholar" ]
Reasoning Models Generate Societies of Thought
Large language models have achieved remarkable capabilities across domains, yet mechanisms underlying sophisticated reasoning remain elusive. Recent reasoning models outperform comparable instruction-tuned models on complex cognitive tasks, attributed to extended computation through longer chains of thought. Here we sh...
[ "Junsol Kim", "Shiyang Lai", "Nino Scherrer", "Blaise Agüera y Arcas", "James Evans" ]
[ "cs.CL", "cs.CY", "cs.LG" ]
[ "Computer Science" ]
2026-01-15T00:00:00
https://arxiv.org/abs/2601.10825
https://arxiv.org/pdf/2601.10825v1
2601.10825
10.48550/arXiv.2601.10825
8
0
false
null
arXiv.org
0.4411
8a1eb36f9c3577289c2bc8e3ebe69c818cdcb82dc5bc02c35926fcbecc4c0746
[ "arxiv", "semantic_scholar" ]
Thinking Long, but Short: Stable Sequential Test-Time Scaling for Large Reasoning Models
Sequential test-time scaling is a promising training-free method to improve large reasoning model accuracy, but as currently implemented, significant limitations have been observed. Inducing models to think for longer can increase their accuracy, but as the length of reasoning is further extended, it has also been show...
[ "Michael R. Metel", "Yufei Cui", "Boxing Chen", "Prasanna Parthasarathi" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2026-01-14T00:00:00
https://arxiv.org/abs/2601.09855
https://arxiv.org/pdf/2601.09855v1
2601.09855
10.48550/arXiv.2601.09855
0
0
false
null
Conference of the European Chapter of the Association for Computational Linguistics
0.44
b13028ebfe89d03d701d050feeaa0b860d52cef8b341dcc4dae9fc163715486b
[ "arxiv", "semantic_scholar" ]
PaCoRe: Learning to Scale Test-Time Compute with Parallel Coordinated Reasoning
We introduce Parallel Coordinated Reasoning (PaCoRe), a training-and-inference framework designed to overcome a central limitation of contemporary language models: their inability to scale test-time compute (TTC) far beyond sequential reasoning under a fixed context window. PaCoRe departs from the traditional sequentia...
[ "Jingcheng Hu", "Yinmin Zhang", "Shijie Shang", "Xiaobo Yang", "Yue Peng", "Zhewei Huang", "Hebin Zhou", "Xin Wu", "Jie Cheng", "Fanqi Wan", "Xiangwen Kong", "Chengyuan Yao", "Kaiwen Yan", "Ailin Huang", "Hongyu Zhou", "Qi Han", "Zheng Ge", "Daxin Jiang", "Xiangyu Zhang", "Heun...
[ "cs.LG" ]
[ "Computer Science" ]
2026-01-09T00:00:00
https://arxiv.org/abs/2601.05593
https://arxiv.org/pdf/2601.05593v1
2601.05593
10.48550/arXiv.2601.05593
13
1
true
null
arXiv.org
0.6711
fb4cdbf8ce418b919226f38ded7a62900158969250e2c70acd2fc7887e99344a
[ "arxiv", "semantic_scholar" ]
TIME: Temporally Intelligent Meta-reasoning Engine for Context-Triggered Explicit Reasoning
Reasoning-oriented language models typically expose explicit reasoning as a long, front-loaded chain of "thinking" tokens before the main output, either always enabled or externally toggled at inference time. Although this can help on arithmetic, coding, and other multi-step tasks, it is costly, weakens claim-level aud...
[ "Susmit Das" ]
[ "cs.LG", "cs.CL" ]
[ "Computer Science" ]
2026-01-08T00:00:00
https://arxiv.org/abs/2601.05300
https://arxiv.org/pdf/2601.05300v2
2601.05300
10.48550/arXiv.2601.05300
1
0
true
https://github.com/The-Coherence-Initiative/TIME
arXiv.org
0.6694
8a95c9986e9bbca77a39675553503c4c8bf830b8e2cf10e168cab0736425efb9
[ "arxiv", "semantic_scholar" ]
ConMax: Confidence-Maximizing Compression for Efficient Chain-of-Thought Reasoning
Recent breakthroughs in Large Reasoning Models (LRMs) have demonstrated that extensive Chain-of-Thought (CoT) generation is critical for enabling intricate cognitive behaviors, such as self-verification and backtracking, to solve complex tasks. However, this capability often leads to ``overthinking'', where models gene...
[ "Minda Hu", "Zexuan Qiu", "Zenan Xu", "Kun Li", "Bo Zhou", "Irwin King" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2026-01-08T00:00:00
https://arxiv.org/abs/2601.04973
https://arxiv.org/pdf/2601.04973v1
2601.04973
10.48550/arXiv.2601.04973
2
0
false
null
arXiv.org
0.4331
4c594dc8cc92db379b309a6bf578ec11c4a79ff3fbbe647b3ba9ac6b821713f0
[ "arxiv", "semantic_scholar" ]
GlimpRouter: Efficient Collaborative Inference by Glimpsing One Token of Thoughts
Large Reasoning Models (LRMs) achieve remarkable performance by explicitly generating multi-step chains of thought, but this capability incurs substantial inference latency and computational cost. Collaborative inference offers a promising solution by selectively allocating work between lightweight and large models, ye...
[ "Wenhao Zeng", "Xuteng Zhang", "Yuling Shi", "Chao Hu", "Yuting Chen", "Beijun Shen", "Xiaodong Gu" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-01-08T00:00:00
https://arxiv.org/abs/2601.05110
https://arxiv.org/pdf/2601.05110v3
2601.05110
10.48550/arXiv.2601.05110
4
1
true
https://github.com/Zengwh02/GlimpRouter
arXiv.org
0.6694
44661096f4a005e552b929966ce423eaa9946148a5f23e65740aeaa2b30d1f8e
[ "arxiv", "semantic_scholar" ]
DiffCoT: Diffusion-styled Chain-of-Thought Reasoning in LLMs
Chain-of-Thought (CoT) reasoning improves multi-step mathematical problem solving in large language models but remains vulnerable to exposure bias and error accumulation, as early mistakes propagate irreversibly through autoregressive decoding. In this work, we propose DiffCoT, a diffusion-styled CoT framework that ref...
[ "Shidong Cao", "Hongzhan Lin", "Yuxuan Gu", "Ziyang Luo", "Jing Ma" ]
[ "cs.CL" ]
[ "Computer Science" ]
2026-01-07T00:00:00
https://arxiv.org/abs/2601.03559
https://arxiv.org/pdf/2601.03559v2
2601.03559
10.48550/arXiv.2601.03559
4
0
false
null
arXiv.org
0.432
ec674373de6cca044356c5606c49665652f874c64dba311c294d0ff99c684886
[ "arxiv", "semantic_scholar" ]
ChaosBench-Logic: A Benchmark for Logical and Symbolic Reasoning on Chaotic Dynamical Systems
Large language models (LLMs) excel at natural language tasks but remain brittle in domains requiring precise logical and symbolic reasoning. Chaotic dynamical systems provide an especially demanding test because chaos is deterministic yet often misinterpreted as randomness or complexity. We introduce ChaosBench-Logic, ...
[ "Noel Thomas" ]
[ "cs.AI", "cs.IT" ]
[ "Computer Science", "Mathematics" ]
2026-01-05T00:00:00
https://arxiv.org/abs/2601.01982
https://arxiv.org/pdf/2601.01982v1
2601.01982
10.48550/arXiv.2601.01982
1
1
true
null
arXiv.org
0.6641
45bdf9998e62051dbc139cc03d39a4c26119cde490b298b55f34ba72cd191ead
[ "arxiv", "semantic_scholar" ]
Falcon-H1R: Pushing the Reasoning Frontiers with a Hybrid Model for Efficient Test-Time Scaling
This work introduces Falcon-H1R, a 7B-parameter reasoning-optimized model that establishes the feasibility of achieving competitive reasoning performance with small language models (SLMs). Falcon-H1R stands out for its parameter efficiency, consistently matching or outperforming SOTA reasoning models that are $2\times$...
[ " Falcon LLM Team", "Iheb Chaabane", "Puneesh Khanna", "Suhail Mohmad", "Slim Frikha", "Shi Hu", "Abdalgader Abubaker", "Reda Alami", "Mikhail Lubinets", "Mohamed El Amine Seddik", "Hakim Hacid" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-01-05T00:00:00
https://arxiv.org/abs/2601.02346
https://arxiv.org/pdf/2601.02346v1
2601.02346
10.48550/arXiv.2601.02346
9
1
false
null
arXiv.org
0.4297
ce781922f11211fb7b5e6e2044c3efd9573752f559cf4cccdf91abb2fad7f8f5
[ "arxiv", "semantic_scholar" ]
Streaming Hallucination Detection in Long Chain-of-Thought Reasoning
Long chain-of-thought (CoT) reasoning improves the performance of large language models, yet hallucinations in such settings often emerge subtly and propagate across reasoning steps. We suggest that hallucination in long CoT reasoning is better understood as an evolving latent state rather than a one-off erroneous even...
[ "Haolang Lu", "Minghui Pan", "Ripeng Li", "Guoshun Nan", "Jialin Zhuang", "Zijie Zhao", "Zhongxiang Sun", "Kun Wang", "Yang Liu" ]
[ "cs.AI" ]
[ "Computer Science" ]
2026-01-05T00:00:00
https://arxiv.org/abs/2601.02170
https://arxiv.org/pdf/2601.02170v1
2601.02170
10.48550/arXiv.2601.02170
2
0
false
null
arXiv.org
0.4297
23cabd5d19680e31a2a0e0b89274b3cc8913bad65762b258476a9882e57df868
[ "arxiv", "semantic_scholar" ]
Watch Wider and Think Deeper: Collaborative Cross-modal Chain-of-Thought for Complex Visual Reasoning
Multi-modal reasoning requires the seamless integration of visual and linguistic cues, yet existing Chain-of-Thought methods suffer from two critical limitations in cross-modal scenarios: (1) over-reliance on single coarse-grained image regions, and (2) semantic fragmentation between successive reasoning steps. To addr...
[ "Wenting Lu", "Didi Zhu", "Tao Shen", "Donglin Zhu", "Ayong Ye", "Chao Wu" ]
[ "cs.CV", "cs.AI" ]
[ "Computer Science" ]
2026-01-04T00:00:00
https://arxiv.org/abs/2601.02422
https://arxiv.org/pdf/2601.02422v1
2601.02422
10.48550/arXiv.2601.02422
0
0
true
https://github.com/deer-echo/CoCoT
arXiv.org
0.6623
224131d3ff11498af1292e3908af516e7eabe6f5db514540051daa99d012b127
[ "arxiv", "semantic_scholar" ]
End-to-End Test-Time Training for Long Context
We formulate long-context language modeling as a problem in continual learning rather than architecture design. Under this formulation, we only use a standard architecture -- a Transformer with sliding-window attention. However, our model continues learning at test time via next-token prediction on the given context, c...
[ "Arnuv Tandon", "Karan Dalal", "Xinhao Li", "Daniel Koceja", "Marcel Rød", "Sam Buchanan", "Xiaolong Wang", "Jure Leskovec", "Sanmi Koyejo", "Tatsunori Hashimoto", "Carlos Guestrin", "Jed McCaleb", "Yejin Choi", "Yu Sun" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-12-29T00:00:00
https://arxiv.org/abs/2512.23675
https://arxiv.org/pdf/2512.23675v2
2512.23675
10.48550/arXiv.2512.23675
26
2
true
https://github.com/test-time-training/e2e
arXiv.org
0.6517
bf151f73a2fbb8388bb83b568828d9ebeeeb5aca3ea2d822ce74f2e8d126aa84
[ "arxiv", "semantic_scholar" ]
Real-Time Service Subscription and Adaptive Offloading Control in Vehicular Edge Computing
Vehicular Edge Computing (VEC) has emerged as a promising paradigm for enhancing the computational efficiency and service quality in intelligent transportation systems by enabling vehicles to wirelessly offload computation-intensive tasks to nearby Roadside Units. However, efficient task offloading and resource allocat...
[ "Chuanchao Gao", "Arvind Easwaran" ]
[ "cs.DC", "cs.DM" ]
[ "Computer Science" ]
2025-12-16T00:00:00
https://arxiv.org/abs/2512.14002
https://arxiv.org/pdf/2512.14002v1
2512.14002
10.1109/RTSS66672.2025.00023
1
0
true
null
IEEE Real-Time Systems Symposium
0.6286
4469ff01c317c8fc8fea5031230d45b312e765a27a264efe4bc145cd34749ab9
[ "arxiv", "semantic_scholar" ]
AutoTool: Dynamic Tool Selection and Integration for Agentic Reasoning
Agentic reinforcement learning has advanced large language models (LLMs) to reason through long chain-of-thought trajectories while interleaving external tool use. Existing approaches assume a fixed inventory of tools, which limits the adaptability of LLM agents to new or evolving toolsets. We present AutoTool, a train...
[ "Jiaru Zou", "Ling Yang", "Yunzhe Qi", "Sirui Chen", "Mengting Ai", "Ke Shen", "Jingrui He", "Mengdi Wang" ]
[ "cs.CL", "cs.LG" ]
[ "Computer Science" ]
2025-12-15T00:00:00
https://arxiv.org/abs/2512.13278
https://arxiv.org/pdf/2512.13278v2
2512.13278
10.48550/arXiv.2512.13278
2
0
false
null
arXiv.org
0.4056
656f2bcc9567e4a9a5f04130cd68f0684f62dbdd5bd4c0c66badee17ac827a9d
[ "arxiv", "semantic_scholar" ]
Limits and Gains of Test-Time Scaling in Vision-Language Reasoning
Test-time scaling (TTS) has emerged as a powerful paradigm for improving the reasoning ability of Large Language Models (LLMs) by allocating additional computation at inference, yet its application to multimodal systems such as Vision-Language Models (VLMs) remains underexplored. In this work, we present a systematic e...
[ "Mohammadjavad Ahmadpour", "Amirmahdi Meighani", "Payam Taebi", "Omid Ghahroodi", "Amirmohammad Izadi", "Mahdieh Soleymani Baghshah" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-12-11T00:00:00
https://arxiv.org/abs/2512.11109
https://arxiv.org/pdf/2512.11109v1
2512.11109
10.48550/arXiv.2512.11109
2
0
true
null
arXiv.org
0.6198
98dfbcc6c11cc2835f5f2cb4bacdf4c2edaf9663f6c2a54293838e0bd1247093
[ "arxiv", "semantic_scholar" ]
Rethinking Chain-of-Thought Reasoning for Videos
Chain-of-thought (CoT) reasoning has been highly successful in solving complex tasks in natural language processing, and recent multimodal large language models (MLLMs) have extended this paradigm to video reasoning. However, these models typically build on lengthy reasoning chains and large numbers of input visual tok...
[ "Yiwu Zhong", "Zi-Yuan Hu", "Yin Li", "Liwei Wang" ]
[ "cs.CV", "cs.AI", "cs.CL", "cs.LG" ]
[ "Computer Science" ]
2025-12-10T00:00:00
https://arxiv.org/abs/2512.09616
https://arxiv.org/pdf/2512.09616v1
2512.09616
10.48550/arXiv.2512.09616
4
0
true
https://github.com/LaVi-Lab/Rethink_CoT_Video
arXiv.org
0.618
fd4b2a00552821ea2f45b886a78cb3604df81ce48dee7ebe6466d86b874a084b
[ "arxiv", "semantic_scholar" ]
MM-CoT:A Benchmark for Probing Visual Chain-of-Thought Reasoning in Multimodal Models
The ability to perform Chain-of-Thought (CoT) reasoning marks a major milestone for multimodal models (MMs), enabling them to solve complex visual reasoning problems. Yet a critical question remains: is such reasoning genuinely grounded in visual evidence and logically coherent? Existing benchmarks emphasize generation...
[ "Jusheng Zhang", "Kaitong Cai", "Xiaoyang Guo", "Sidi Liu", "Qinhan Lv", "Ruiqi Chen", "Jing Yang", "Yijia Fan", "Xiaofei Sun", "Jian Wang", "Ziliang Chen", "Liang Lin", "Keze Wang" ]
[ "cs.CV", "cs.AI" ]
[ "Computer Science" ]
2025-12-09T00:00:00
https://arxiv.org/abs/2512.08228
https://arxiv.org/pdf/2512.08228v1
2512.08228
10.48550/arXiv.2512.08228
35
2
false
null
arXiv.org
0.3987
509bc2159dbca241853f6087b28ff21c4f9180ce2ca0cae36295e8c732aedc71
[ "arxiv", "semantic_scholar" ]
RoBoN: Routed Online Best-of-n for Test-Time Scaling with Multiple LLMs
Best-of-$n$ is a widely used test-time scaling approach for LLM inference. Yet despite evidence that LLMs exhibit complementary strengths across tasks, traditionally best-of-$n$ relies on a single model to generate responses. We propose RoBoN (Routed Online Best-of-$n$), a sequential multi-LLM alternative to the prevai...
[ "Jonathan Geuter", "Gregor Kornhardt" ]
[ "cs.LG", "cs.AI" ]
[ "Computer Science" ]
2025-12-05T00:00:00
https://arxiv.org/abs/2512.05542
https://arxiv.org/pdf/2512.05542v1
2512.05542
10.48550/arXiv.2512.05542
0
0
false
null
arXiv.org
0.3942
4620bc0b817a2ddd6d16265f9a890ff30ac5e0d04016cd52b161b49fd20975c0
[ "arxiv", "semantic_scholar" ]
Probing the effectiveness of World Models for Spatial Reasoning through Test-time Scaling
Vision-Language Models (VLMs) remain limited in spatial reasoning tasks that require multi-view understanding and embodied perspective shifts. Recent approaches such as MindJourney attempt to mitigate this gap through test-time scaling where a world model imagines action-conditioned trajectories and a heuristic verifie...
[ "Saurav Jha", "M. Jehanzeb Mirza", "Wei Lin", "Shiqi Yang", "Sarath Chandar" ]
[ "cs.CV", "cs.AI" ]
[ "Computer Science" ]
2025-12-05T00:00:00
https://arxiv.org/abs/2512.05809
https://arxiv.org/pdf/2512.05809v1
2512.05809
10.48550/arXiv.2512.05809
0
0
true
https://github.com/chandar-lab/visa-for-mindjourney
arXiv.org
0.6092
2250cb6b6ce29df096a5e16d34cbf68d8bf772f263ef5acc0628dc954adb8900
[ "arxiv", "semantic_scholar" ]
Visual Reasoning Tracer: Object-Level Grounded Reasoning Benchmark
Recent advances in Multimodal Large Language Models (MLLMs) have significantly improved performance on tasks such as visual grounding and visual question answering. However, the reasoning processes of these models remain largely opaque; they typically output only final predictions without revealing the intermediate ste...
[ "Haobo Yuan", "Yueyi Sun", "Yanwei Li", "Tao Zhang", "Xueqing Deng", "Henghui Ding", "Lu Qi", "Anran Wang", "Xiangtai Li", "Ming-Hsuan Yang" ]
[ "cs.CV" ]
[ "Computer Science" ]
2025-12-04T00:00:00
https://arxiv.org/abs/2512.05091
https://arxiv.org/pdf/2512.05091v1
2512.05091
10.48550/arXiv.2512.05091
8
0
false
null
arXiv.org
0.393
5d284a5556216242533927d57448e80ed7ed9a88528748c6052d69252b2686b5
[ "arxiv", "semantic_scholar" ]
The Art of Scaling Test-Time Compute for Large Language Models
Test-time scaling (TTS) -- the dynamic allocation of compute during inference -- is a promising direction for improving reasoning in large language models (LLMs). However, a systematic comparison of well-known TTS strategies under identical conditions is missing, and the influence of model type and problem difficulty o...
[ "Aradhye Agarwal", "Ayan Sengupta", "Tanmoy Chakraborty" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-12-01T00:00:00
https://arxiv.org/abs/2512.02008
https://arxiv.org/pdf/2512.02008v1
2512.02008
10.48550/arXiv.2512.02008
9
0
true
null
arXiv.org
0.6021
eb574edb918b77f9378bdbdd439900335030631eb5a00547a922fbf993c3652e
[ "arxiv", "semantic_scholar" ]
Reasoning Under Pressure: How do Training Incentives Influence Chain-of-Thought Monitorability?
AI systems that output their reasoning in natural language offer an opportunity for safety -- we can \emph{monitor} their chain of thought (CoT) for undesirable reasoning, such as the pursuit of harmful objectives. However, the extent to which CoT faithfully reflects the underlying reasoning process, and hence the exte...
[ "Matt MacDermott", "Qiyao Wei", "Rada Djoneva", "Francis Rhys Ward" ]
[ "cs.AI", "cs.CR" ]
[ "Computer Science" ]
2025-11-28T00:00:00
https://arxiv.org/abs/2512.00218
https://arxiv.org/pdf/2512.00218v2
2512.00218
10.48550/arXiv.2512.00218
2
0
true
https://github.com/QiyaoWei/reasoning-under-pressure
arXiv.org
0.5968
e811c855a83df33f20c6f5fce8cf85709a576c903e589562e2ab1de69e7494e1
[ "arxiv", "semantic_scholar" ]
Focused Chain-of-Thought: Efficient LLM Reasoning via Structured Input Information
Recent large language models achieve strong reasoning performance by generating detailed chain-of-thought traces, but this often leads to excessive token use and high inference latency. Existing efficiency approaches typically focus on model-centric interventions, such as reinforcement learning or supervised fine-tunin...
[ "Lukas Struppek", "Dominik Hintersdorf", "Hannah Struppek", "Daniel Neider", "Kristian Kersting" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-11-27T00:00:00
https://arxiv.org/abs/2511.22176
https://arxiv.org/pdf/2511.22176v1
2511.22176
10.48550/arXiv.2511.22176
1
0
false
null
arXiv.org
0.385
715075618f157752c13480f07e6136babcff9f8d7fa91970350a6e6f85ffb915
[ "arxiv", "semantic_scholar" ]
VICoT-Agent: A Vision-Interleaved Chain-of-Thought Framework for Interpretable Multimodal Reasoning and Scalable Remote Sensing Analysis
The current remote sensing image analysis task is increasingly evolving from traditional object recognition to complex intelligence reasoning, which places higher requirements on the model's reasoning ability and the flexibility of tool invocation. To this end, we propose a new multimodal agent framework, Vision-Interl...
[ "Chujie Wang", "Zhiyuan Luo", "Ruiqi Liu", "Can Ran", "Shenghua Fan", "Xi Chen", "Chu He" ]
[ "cs.AI", "cs.MA" ]
[ "Computer Science" ]
2025-11-25T00:00:00
https://arxiv.org/abs/2511.20085
https://arxiv.org/pdf/2511.20085v3
2511.20085
10.48550/arXiv.2511.20085
3
0
false
null
arXiv.org
0.3827
575d9c447db11af25482361a575f5b4447ac8e603de12469b0e54a190382897e
[ "arxiv", "semantic_scholar" ]
VisReason: A Large-Scale Dataset for Visual Chain-of-Thought Reasoning
Chain-of-Thought (CoT) prompting has proven remarkably effective for eliciting complex reasoning in large language models (LLMs). Yet, its potential in multimodal large language models (MLLMs) remains largely untapped, hindered by the absence of large-scale datasets that capture the rich, spatially grounded reasoning i...
[ "Lingxiao Li", "Yifan Wang", "Xinyan Gao", "Chen Tang", "Xiangyu Yue", "Chenyu You" ]
[ "cs.CV", "cs.LG" ]
[ "Computer Science" ]
2025-11-21T00:00:00
https://arxiv.org/abs/2511.17731
https://arxiv.org/pdf/2511.17731v1
2511.17731
10.48550/arXiv.2511.17731
3
0
false
null
arXiv.org
0.3781
1dea087c7c6c29001a9c47ba7731720b9e3742da5d7ed80bfb66362757bb74ba
[ "arxiv", "semantic_scholar" ]
Geospatial Chain of Thought Reasoning for Enhanced Visual Question Answering on Satellite Imagery
Geospatial chain of thought (CoT) reasoning is essential for advancing Visual Question Answering (VQA) on satellite imagery, particularly in climate related applications such as disaster monitoring, infrastructure risk assessment, urban resilience planning, and policy support. Existing VQA models enable scalable interp...
[ "Shambhavi Shanker", "Manikandan Padmanaban", "Jagabondhu Hazra" ]
[ "cs.CV" ]
[ "Computer Science" ]
2025-11-14T00:00:00
https://arxiv.org/abs/2511.11198
https://arxiv.org/pdf/2511.11198v1
2511.11198
10.48550/arXiv.2511.11198
1
0
false
null
arXiv.org
0.3701
1ecbc5b401e14e6d91559b47aef01a8e6e412cd445abea095e94352ea3460aae
[ "arxiv", "semantic_scholar" ]
BadThink: Triggered Overthinking Attacks on Chain-of-Thought Reasoning in Large Language Models
Recent advances in Chain-of-Thought (CoT) prompting have substantially improved the reasoning capabilities of large language models (LLMs), but have also introduced their computational efficiency as a new attack surface. In this paper, we propose BadThink, the first backdoor attack designed to deliberately induce "over...
[ "Shuaitong Liu", "Renjue Li", "Lijia Yu", "Lijun Zhang", "Zhiming Liu", "Gaojie Jin" ]
[ "cs.CR", "cs.AI" ]
[ "Computer Science" ]
2025-11-13T00:00:00
https://arxiv.org/abs/2511.10714
https://arxiv.org/pdf/2511.10714v1
2511.10714
10.48550/arXiv.2511.10714
4
0
false
null
arXiv.org
0.369
b672b8b26bd81c6c57e85ac7b4c22a54d37f4cb5fb4d218080467e70bbe4af32
[ "arxiv", "semantic_scholar" ]
Benchmarking Multi-Step Legal Reasoning and Analyzing Chain-of-Thought Effects in Large Language Models
Large language models (LLMs) have demonstrated strong reasoning abilities across specialized domains, motivating research into their application to legal reasoning. However, existing legal benchmarks often conflate factual recall with genuine inference, fragment the reasoning process, and overlook the quality of reason...
[ "Wenhan Yu", "Xinbo Lin", "Lanxin Ni", "Jinhua Cheng", "Lei Sha" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-11-11T00:00:00
https://arxiv.org/abs/2511.07979
https://arxiv.org/pdf/2511.07979v3
2511.07979
10.48550/arXiv.2511.07979
3
0
true
https://github.com/yuwenhan07/MSLR-Bench
arXiv.org
0.5667
a9930d9c773c75d3e041b198b78fb77ecb881476730de75cc0b70a585e64d58f
[ "arxiv", "semantic_scholar" ]
Think Consistently, Reason Efficiently: Energy-Based Calibration for Implicit Chain-of-Thought
Large Language Models (LLMs) have demonstrated strong reasoning capabilities through \emph{Chain-of-Thought} (CoT) prompting, which enables step-by-step intermediate reasoning. However, explicit CoT methods rely on discrete token-level reasoning processes that are prone to error propagation and limited by vocabulary ex...
[ "Zhikang Chen", "Sen Cui", "Deheng Ye", "Yu Zhang", "Yatao Bian", "Tingting Zhu" ]
[ "cs.CL", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-11-10T00:00:00
https://arxiv.org/abs/2511.07124
https://arxiv.org/pdf/2511.07124v1
2511.07124
10.48550/arXiv.2511.07124
1
1
false
null
arXiv.org
0.3655
99df5803636363c6d06362f52a2f7f3ae6f47badf9fc980ff7b18b44784f6c0e
[ "arxiv", "semantic_scholar" ]
Chain-of-Thought as a Lens: Evaluating Structured Reasoning Alignment between Human Preferences and Large Language Models
This paper primarily demonstrates a method to quantitatively assess the alignment between multi-step, structured reasoning in large language models and human preferences. We introduce the Alignment Score, a semantic-level metric that compares a model-produced chain of thought traces with a human-preferred reference by ...
[ "Boxuan Wang", "Zhuoyun Li", "Xinmiao Huang", "Xiaowei Huang", "Yi Dong" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-11-09T00:00:00
https://arxiv.org/abs/2511.06168
https://arxiv.org/pdf/2511.06168v3
2511.06168
null
2
0
true
null
null
0.4306
d73f6edce588bfedb77dc501bc5ed57ba614ab99c698ecfc8bfaa418405e57c2
[ "arxiv", "semantic_scholar" ]
Logit-Entropy Adaptive Stopping Heuristic for Efficient Chain-of-Thought Reasoning
Chain-of-Thought (CoT) prompting is a key technique for enabling complex reasoning in large language models. However, generating full, fixed-length rationales is computationally wasteful, inflating both token usage and latency. We introduce LEASH: Logit-Entropy Adaptive Stopping Heuristic, a training-free decoding algo...
[ "Mohammad Atif Quamar", "Mohammad Areeb" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-11-06T00:00:00
https://arxiv.org/abs/2511.04654
https://arxiv.org/pdf/2511.04654v1
2511.04654
10.48550/arXiv.2511.04654
1
0
false
null
arXiv.org
0.3609
2a2312a540f43afe3f9c1138eaaf927dc6d16347043f325043422e0ea8873c64
[ "arxiv", "semantic_scholar" ]
Efficient Reasoning via Thought-Training and Thought-Free Inference
Recent advances in large language models (LLMs) have leveraged explicit Chain-of-Thought (CoT) prompting to improve reasoning accuracy. However, most existing methods primarily focus on compressing verbose reasoning outputs. These Long-to-Short transformations aim to improve efficiency, but require a large amount of sh...
[ "Canhui Wu", "Qiong Cao", "Chao Xue", "Wei Xi", "Xiaodong He" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-11-05T00:00:00
https://arxiv.org/abs/2511.03408
https://arxiv.org/pdf/2511.03408v3
2511.03408
10.48550/arXiv.2511.03408
2
0
false
null
arXiv.org
0.3598
1b36ce09bbcb9e3deb8c1a069f592058c6342aab58eb4f068afa1283e4d0bfbd
[ "arxiv", "semantic_scholar" ]
Optimizing Reasoning Efficiency through Prompt Difficulty Prediction
Reasoning language models perform well on complex tasks but are costly to deploy due to their size and long reasoning traces. We propose a routing approach that assigns each problem to the smallest model likely to solve it, reducing compute without sacrificing accuracy. Using intermediate representations from s1.1-32B,...
[ "Bo Zhao", "Berkcan Kapusuzoglu", "Kartik Balasubramaniam", "Sambit Sahu", "Supriyo Chakraborty", "Genta Indra Winata" ]
[ "cs.LG", "cs.AI" ]
[ "Computer Science" ]
2025-11-05T00:00:00
https://arxiv.org/abs/2511.03808
https://arxiv.org/pdf/2511.03808v1
2511.03808
10.48550/arXiv.2511.03808
2
0
false
null
arXiv.org
0.3598
f3330111932fc555d2499736590d6797f5e6ae0aa34e61d6b486c5280d494264
[ "arxiv", "semantic_scholar" ]
Re-FORC: Adaptive Reward Prediction for Efficient Chain-of-Thought Reasoning
We propose Re-FORC, an adaptive reward prediction method that, given a context, enables prediction of the expected future rewards as a function of the number of future thinking tokens. Re-FORC trains a lightweight adapter on reasoning models, demonstrating improved prediction with longer reasoning and larger models. Re...
[ "Renos Zabounidis", "Aditya Golatkar", "Michael Kleinman", "Alessandro Achille", "Wei Xia", "Stefano Soatto" ]
[ "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-11-03T00:00:00
https://arxiv.org/abs/2511.02130
https://arxiv.org/pdf/2511.02130v1
2511.02130
10.48550/arXiv.2511.02130
2
0
false
null
arXiv.org
0.3575
15e1a44ac5d4b43b21148c396ea9285fe825ac9ca10d8986214b5a3d2a222a5e
[ "arxiv", "semantic_scholar" ]
Scaling Graph Chain-of-Thought Reasoning: A Multi-Agent Framework with Efficient LLM Serving
Graph Chain-of-Thought (Graph-CoT) enables large language models (LLMs) to perform step-by-step reasoning over graph-structured knowledge, but existing pipelines suffer from low accuracy, excessive token usage, high latency, and low throughput due to single-agent monolithic prompts, repeated context re-encoding, and in...
[ "Chengying Huan", "Ziheng Meng", "Yongchao Liu", "Zhengyi Yang", "Yun Zhu", "Yue Yun", "Shipeng Li", "Rong Gu", "Xiabao Wu", "Haitao Zhang", "Chuntao Hong", "Shaonan Ma", "Guihai Chen", "Chen Tian" ]
[ "cs.LG", "cs.AI" ]
[ "Computer Science" ]
2025-11-03T00:00:00
https://arxiv.org/abs/2511.01633
https://arxiv.org/pdf/2511.01633v1
2511.01633
10.48550/arXiv.2511.01633
5
0
false
null
arXiv.org
0.3575
66994dea891f2af79299a9cd30ac1f0e27f41c6016318ebe2ec1ee0ba4d58393
[ "arxiv", "semantic_scholar" ]
Inference-Time Chain-of-Thought Pruning with Latent Informativeness Signals
Large language models (LLMs) improve reasoning accuracy when generating multiple candidate solutions at test time, but standard methods like Best-of-N (BoN) incur high computational cost by fully generating all branches. Self-Truncation Best-of-N (ST-BoN) mitigates this by truncating unpromising paths early, but its re...
[ "Sophie Li", "Nicholas Huang", "Nayan Saxena", "Nina Luo", "Vincent Lin", "Kevin Zhu", "Sunishchal Dev" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-11-01T00:00:00
https://arxiv.org/abs/2511.00699
https://arxiv.org/pdf/2511.00699v2
2511.00699
10.48550/arXiv.2511.00699
1
0
false
null
arXiv.org
0.3552
96bd7c39750feff44218d5b86bdf427badf16ed3e98659186efbc5d6dab3b044
[ "arxiv", "semantic_scholar" ]
Reasoning Models Sometimes Output Illegible Chains of Thought
Language models trained via outcome-based reinforcement learning (RL) to reason using chain-of-thought (CoT) have shown remarkable performance. Monitoring such a model's CoT may allow us to understand its intentions and detect potential malicious behavior. However, to be effective, this requires that CoTs are legible a...
[ "Arun Jose" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-10-31T00:00:00
https://arxiv.org/abs/2510.27338
https://arxiv.org/pdf/2510.27338v1
2510.27338
10.48550/arXiv.2510.27338
7
1
false
null
arXiv.org
0.3541
21c0fc8d983ae80039f3fe6e382247070c09e31e8c5fac33da26d29a98ba213a
[ "arxiv", "semantic_scholar" ]
Diffuse Thinking: Exploring Diffusion Language Models as Efficient Thought Proposers for Reasoning
In recent years, large language models (LLMs) have witnessed remarkable advancements, with the test-time scaling law consistently enhancing the reasoning capabilities. Through systematic evaluation and exploration of a diverse spectrum of intermediate thoughts, LLMs demonstrate the potential to generate deliberate reas...
[ "Chenyang Shao", "Sijian Ren", "Fengli Xu", "Yong Li" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-31T00:00:00
https://arxiv.org/abs/2510.27469
https://arxiv.org/pdf/2510.27469v1
2510.27469
10.48550/arXiv.2510.27469
4
0
true
null
arXiv.org
0.5472
7da963ec9bb8667fa42b94ccf8c4055a5d523295cf0aff8b2eb136e9fb952e4f
[ "arxiv", "semantic_scholar" ]
ThinkMorph: Emergent Properties in Multimodal Interleaved Chain-of-Thought Reasoning
Multimodal reasoning requires iterative coordination between language and vision, yet it remains unclear what constitutes a meaningful interleaved chain of thought. We posit that text and image thoughts should function as complementary rather than isomorphic modalities that mutually advance reasoning. Guided by this pr...
[ "Jiawei Gu", "Yunzhuo Hao", "Huichen Will Wang", "Linjie Li", "Michael Qizhe Shieh", "Yejin Choi", "Ranjay Krishna", "Yu Cheng" ]
[ "cs.CV" ]
[ "Computer Science" ]
2025-10-30T00:00:00
https://arxiv.org/abs/2510.27492
https://arxiv.org/pdf/2510.27492v3
2510.27492
10.48550/arXiv.2510.27492
31
5
false
null
arXiv.org
0.3891
9f732b17e9601e8bb0e5408f32f7dd29ccb006f10f8113d75dd2b9cfe3a2dad8
[ "arxiv", "semantic_scholar" ]
Chopping Trees: Semantic Similarity Based Dynamic Pruning for Tree-of-Thought Reasoning
Tree-of-Thought (ToT) reasoning boosts the problem-solving abilities of Large Language Models (LLMs) but is computationally expensive due to semantic redundancy, where distinct branches explore equivalent reasoning paths. We introduce Semantic Similarity-Based Dynamic Pruning (SSDP), a lightweight method that, to the b...
[ "Joongho Kim", "Xirui Huang", "Zarreen Reza", "Gabriel Grand" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-10-30T00:00:00
https://arxiv.org/abs/2511.08595
https://arxiv.org/pdf/2511.08595v2
2511.08595
10.48550/arXiv.2511.08595
2
1
true
https://github.com/kimjoonghokim/SSDP
arXiv.org
0.5454
797449e3ab2288258e52b18a384461dae94dd7157c57fffe226140bae5a2f21b
[ "arxiv", "semantic_scholar" ]
Latent Chain-of-Thought for Visual Reasoning
Chain-of-thought (CoT) reasoning is critical for improving the interpretability and reliability of Large Vision-Language Models (LVLMs). However, existing training algorithms such as SFT, PPO, and GRPO may not generalize well across unseen reasoning tasks and heavily rely on a biased reward model. To address this chall...
[ "Guohao Sun", "Hang Hua", "Jian Wang", "Jiebo Luo", "Sohail Dianat", "Majid Rabbani", "Raghuveer Rao", "Zhiqiang Tao" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science", "Medicine" ]
2025-10-27T00:00:00
https://arxiv.org/abs/2510.23925
https://arxiv.org/pdf/2510.23925v2
2510.23925
10.48550/arXiv.2510.23925
24
0
false
null
Advances in Neural Information Processing Systems
0.3495
b73027b0b25e280060d8002b21f0af09cadc30ce78b06418dc78c40d7845348f
[ "arxiv", "semantic_scholar" ]
When Fewer Layers Break More Chains: Layer Pruning Harms Test-Time Scaling in LLMs
Layer pruning has emerged as a widely adopted technique for improving the efficiency of large language models (LLMs). Although existing methods demonstrate strong performance retention on general knowledge tasks, their effect on long-chain reasoning, a more brittle yet crucial capability, remains largely unexplored. In...
[ "Keyu Wang", "Tian Lyu", "Guinan Su", "Jonas Geiping", "Lu Yin", "Marco Canini", "Shiwei Liu" ]
[ "cs.LG", "cs.AI" ]
[ "Computer Science" ]
2025-10-25T00:00:00
https://arxiv.org/abs/2510.22228
https://arxiv.org/pdf/2510.22228v1
2510.22228
10.48550/arXiv.2510.22228
5
2
true
https://github.com/keyu-wang-2002/Layer-Pruning-Harms-Inference-Scaling}{https://github.com/keyu-wang-2002/Layer-Pruning-Harms-Inference-Scaling}
arXiv.org
0.5366
206c0fec87578f8a8ff5b2fd5f7a8db81f5d11af5e9d515ac327372d11134c26
[ "arxiv", "semantic_scholar" ]
The Virtues of Brevity: Avoid Overthinking in Parallel Test-Time Reasoning
Reasoning models represent a significant advance in LLM capabilities, particularly for complex reasoning tasks such as mathematics and coding. Previous studies confirm that parallel test-time compute-sampling multiple solutions and selecting the best one-can further enhance the predictive performance of LLMs. However, ...
[ "Raul Cavalcante Dinardi", "Bruno Yamamoto", "Anna Helena Reali Costa", "Artur Jordao" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-10-24T00:00:00
https://arxiv.org/abs/2510.21067
https://arxiv.org/pdf/2510.21067v1
2510.21067
10.48550/arXiv.2510.21067
0
0
false
null
arXiv.org
0.346
05a11242200d29acd433c401cb75ac372b588c5dc87b4f08fc129f61e72dc451
[ "arxiv", "semantic_scholar" ]
The Zero-Step Thinking: An Empirical Study of Mode Selection as Harder Early Exit in Reasoning Models
Reasoning models have demonstrated exceptional performance in tasks such as mathematics and logical reasoning, primarily due to their ability to engage in step-by-step thinking during the reasoning process. However, this often leads to overthinking, resulting in unnecessary computational overhead. To address this issue...
[ "Yuqiao Tan", "Shizhu He", "Kang Liu", "Jun Zhao" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2025-10-22T00:00:00
https://arxiv.org/abs/2510.19176
https://arxiv.org/pdf/2510.19176v1
2510.19176
10.48550/arXiv.2510.19176
4
0
true
https://github.com/Trae1ounG/Zero_Step_Thinking
arXiv.org
0.5312
1a2f955b846e5935392bd44f1a17ff0179cac30fe05630a0a70631d3de92a672
[ "arxiv", "semantic_scholar" ]
SmartSwitch: Advancing LLM Reasoning by Overcoming Underthinking via Promoting Deeper Thought Exploration
The long chain-of-thought (LongCoT) capability is central to the recent breakthroughs achieved by large language models in complex reasoning tasks. However, the accompanying issue of ''underthinking'', where models exhibit shallow reasoning by frequently switching thoughts without sufficient exploration, limits both pe...
[ "Xichen Zhang", "Sitong Wu", "Haoru Tan", "Shaozuo Yu", "Yinghao Zhu", "Ziyi He", "Jiaya Jia" ]
[ "cs.CL", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-10-22T00:00:00
https://arxiv.org/abs/2510.19767
https://arxiv.org/pdf/2510.19767v1
2510.19767
10.48550/arXiv.2510.19767
1
0
true
https://github.com/dvlab-research/SmartSwitch
arXiv.org
0.5312
de6ce1c329fd310a9f590c5d8bf687e235600ff832627449084adf14b4c5e2ac
[ "arxiv", "semantic_scholar" ]
Can Reasoning Models Obfuscate Reasoning? Stress-Testing Chain-of-Thought Monitorability
Recent findings suggest that misaligned models may exhibit deceptive behavior, raising concerns about output trustworthiness. Chain-of-thought (CoT) is a promising tool for alignment monitoring: when models articulate their reasoning faithfully, monitors can detect and mitigate harmful behaviors before undesirable outc...
[ "Artur Zolkowski", "Wen Xing", "David Lindner", "Florian Tramèr", "Erik Jenner" ]
[ "cs.CR", "cs.AI" ]
[ "Computer Science" ]
2025-10-21T00:00:00
https://arxiv.org/abs/2510.19851
https://arxiv.org/pdf/2510.19851v1
2510.19851
10.48550/arXiv.2510.19851
10
1
false
null
arXiv.org
0.3426
037446ba4b5209794e8852cd6f323ba33ef2b3663a358ac8f53c9c5b288feef3
[ "arxiv", "semantic_scholar" ]
Fine-Tuned Thoughts: Leveraging Chain-of-Thought Reasoning for Industrial Asset Health Monitoring
Small Language Models (SLMs) are becoming increasingly popular in specialized fields, such as industrial applications, due to their efficiency, lower computational requirements, and ability to be fine-tuned for domain-specific tasks, enabling accurate and cost-effective solutions. However, performing complex reasoning ...
[ "Shuxin Lin", "Dhaval Patel", "Christodoulos Constantinides" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-10-21T00:00:00
https://arxiv.org/abs/2510.18817
https://arxiv.org/pdf/2510.18817v1
2510.18817
10.18653/v1/2025.findings-emnlp.1126
2
0
true
https://github.com/IBM/FailureSensorIQ
Conference on Empirical Methods in Natural Language Processing
0.5295
351218de380050c6bf0a44289042e0f37cca3bdcd89953e6921724648a84f533
[ "arxiv", "semantic_scholar" ]
Inference-Time Compute Scaling For Flow Matching
Allocating extra computation at inference time has recently improved sample quality in large language models and diffusion-based image generation. In parallel, Flow Matching (FM) has gained traction in language, vision, and scientific domains, but inference-time scaling methods for it remain under-explored. Concurrentl...
[ "Adam Stecklov", "Noah El Rimawi-Fine", "Mathieu Blanchette" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-10-20T00:00:00
https://arxiv.org/abs/2510.17786
https://arxiv.org/pdf/2510.17786v1
2510.17786
10.48550/arXiv.2510.17786
1
0
false
null
arXiv.org
0.3415
c9e6e4e90a270cad29999fc2bc013f88efbfb7c724f9b953ef24e2b8e559f451
[ "arxiv", "semantic_scholar" ]
Big Reasoning with Small Models: Instruction Retrieval at Inference Time
Small language models (SLMs) enable low-cost, private, on-device inference, but they often fail on problems that require specialized domain knowledge or multi-step reasoning. Existing approaches for improving reasoning either rely on scale (e.g., chain-of-thought prompting), require task-specific training that limits r...
[ "Kenan Alkiek", "David Jurgens", "Vinod Vydiswaran" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-10-15T00:00:00
https://arxiv.org/abs/2510.13935
https://arxiv.org/pdf/2510.13935v2
2510.13935
10.48550/arXiv.2510.13935
0
0
false
null
arXiv.org
0.3357
9664d40ccc24dfdaa486e2ab999de9f732249458b3fa043f05babf3027ab24e9
[ "arxiv", "semantic_scholar" ]
Putting on the Thinking Hats: A Survey on Chain of Thought Fine-tuning from the Perspective of Human Reasoning Mechanism
Chain of thought (CoT) fine-tuning aims to endow large language models (LLMs) with reasoning capabilities by training them on curated reasoning traces. It leverages both supervised and reinforced fine-tuning to cultivate human-like reasoning skills in LLMs, including detailed planning, divergent thinking, intuitive jud...
[ "Xiaoshu Chen", "Sihang Zhou", "Ke Liang", "Duanyang Yuan", "Haoyuan Chen", "Xiaoyu Sun", "Lingyuan Meng", "Xinwang Liu" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-15T00:00:00
https://arxiv.org/abs/2510.13170
https://arxiv.org/pdf/2510.13170v2
2510.13170
10.48550/arXiv.2510.13170
4
0
true
https://github.com/AI-Chen/Awesome-CoT-Finetuning}
arXiv.org
0.5189
1696ed340c4f9f1aececed948345da02b6ee41d12aca8b8f7abb17d0233d86fb
[ "arxiv", "semantic_scholar" ]
Refine Thought: A Test-Time Inference Method for Embedding Model Reasoning
We propose RT (Refine Thought), a method that can enhance the semantic rea-soning ability of text embedding models. The method obtains the final semanticrepresentation by running multiple forward passes of the text embedding model.Experiments show that RT achieves significant improvements on semantic reason-ing tasks i...
[ "Guangzhi Wang", "Kai Li", "Yinghao Jiao", "Zhi Liu" ]
[ "cs.CL", "cs.AI", "cs.IR" ]
[ "Computer Science" ]
2025-10-14T00:00:00
https://arxiv.org/abs/2511.13726
https://arxiv.org/pdf/2511.13726v1
2511.13726
10.48550/arXiv.2511.13726
2
0
false
null
arXiv.org
0.3346
f1a79737ccafb295191359ef98c49b4ba1445a95b7a78f192167c4254586bb4f
[ "arxiv", "semantic_scholar" ]
HiCoTraj:Zero-Shot Demographic Reasoning via Hierarchical Chain-of-Thought Prompting from Trajectory
Inferring demographic attributes such as age, sex, or income level from human mobility patterns enables critical applications such as targeted public health interventions, equitable urban planning, and personalized transportation services. Existing mobility-based demographic inference studies heavily rely on large-scal...
[ "Junyi Xie", "Yuankun Jiao", "Jina Kim", "Yao-Yi Chiang", "Lingyi Zhao", "Khurram Shafique" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-10-14T00:00:00
https://arxiv.org/abs/2510.12067
https://arxiv.org/pdf/2510.12067v1
2510.12067
10.1145/3764915.3770723
0
0
false
null
null
0.2129
7530d3b1852b100f13ee058c6f3769bbc01b3c79a03b9993aa82dd0f1e3a1312
[ "arxiv", "semantic_scholar" ]
Towards Inference-time Scaling for Continuous Space Reasoning
Inference-time scaling through multiple sample generation in combination with Process- or Outcome-Reward Model (PRM or ORM) re-ranking has proven effective for text-based reasoning in large language models. This paper investigates whether such established techniques can be successfully adapted to reasoning in the conti...
[ "Minghan Wang", "Thuy-Trang Vu", "Ehsan Shareghi", "Gholamreza Haffari" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-14T00:00:00
https://arxiv.org/abs/2510.12167
https://arxiv.org/pdf/2510.12167v2
2510.12167
10.48550/arXiv.2510.12167
3
0
false
null
arXiv.org
0.3346
4759d56853651779c5e8300c9c084422ec588d1f3cf91ef6725d72a84f2b077a
[ "arxiv", "semantic_scholar" ]
Enhancing Long Chain-of-Thought Reasoning through Multi-Path Plan Aggregation
Inference-time scaling enhances the reasoning ability of a language model (LM) by extending its chain-of-thought (CoT). However, existing approaches typically generate the entire reasoning chain in a single forward pass, which often leads to CoT derailment, i.e., the reasoning trajectory drifting off course due to comp...
[ "Siheng Xiong", "Ali Payani", "Faramarz Fekri" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-13T00:00:00
https://arxiv.org/abs/2510.11620
https://arxiv.org/pdf/2510.11620v2
2510.11620
10.48550/arXiv.2510.11620
6
0
false
null
arXiv.org
0.3334
663e58bdb0545fc8a5aee57462cca199ff52402002263a2ea7622b738fb579be
[ "arxiv", "semantic_scholar" ]
CoPRS: Learning Positional Prior from Chain-of-Thought for Reasoning Segmentation
Existing works on reasoning segmentation either connect hidden features from a language model directly to a mask decoder or represent positions in text, which limits interpretability and semantic detail. To solve this, we present CoPRS, a Multi-modal Chain-of-Thought (MCoT)-based positional perception model that bridge...
[ "Zhenyu Lu", "Liupeng Li", "Jinpeng Wang", "Yan Feng", "Bin Chen", "Ke Chen", "Yaowei Wang" ]
[ "cs.CV", "cs.MM" ]
[ "Computer Science" ]
2025-10-13T00:00:00
https://arxiv.org/abs/2510.11173
https://arxiv.org/pdf/2510.11173v3
2510.11173
10.48550/arXiv.2510.11173
7
1
true
https://github.com/ZhenyuLU-Heliodore/CoPRS
arXiv.org
0.5153
e5bec2a2dc68878c9cc3c338d03f3f18a16b473cdd9ea51d90707b2acb36539d
[ "arxiv", "semantic_scholar" ]
Review of Inference-Time Scaling Strategies: Reasoning, Search and RAG
The performance gains of LLMs have historically been driven by scaling up model size and training data. However, the rapidly diminishing availability of high-quality training data is introducing a fundamental bottleneck, shifting the focus of research toward inference-time scaling. This paradigm uses additional computa...
[ "Zhichao Wang", "Cheng Wan", "Dong Nie" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-12T00:00:00
https://arxiv.org/abs/2510.10787
https://arxiv.org/pdf/2510.10787v1
2510.10787
10.48550/arXiv.2510.10787
0
0
false
null
arXiv.org
0.3323
0a3cf254180c3e57ee98db84e7911c033f330a3c7df8ce2c40437e6ac8c1bb2d
[ "arxiv", "semantic_scholar" ]
Adaptive Dual Reasoner: Large Reasoning Models Can Think Efficiently by Hybrid Reasoning
Although Long Reasoning Models (LRMs) have achieved superior performance on various reasoning scenarios, they often suffer from increased computational costs and inference latency caused by overthinking. To address these limitations, we propose Adaptive Dual Reasoner, which supports two reasoning modes: fast thinking a...
[ "Yujian Zhang", "Keyu Chen", "Zhifeng Shen", "Ruizhi Qiao", "Xing Sun" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-10-11T00:00:00
https://arxiv.org/abs/2510.10207
https://arxiv.org/pdf/2510.10207v2
2510.10207
10.48550/arXiv.2510.10207
2
1
false
null
arXiv.org
0.3311
1c7e9d3af53639fd5e068b5a1da5439155dae0dbb04749b6e31e71d1cbb503c3
[ "arxiv", "semantic_scholar" ]
MatryoshkaThinking: Recursive Test-Time Scaling Enables Efficient Reasoning
Test-time scaling has emerged as a promising paradigm in language modeling, wherein additional computational resources are allocated during inference to enhance model performance. Recent approaches, such as DeepConf, have demonstrated the efficacy of this strategy, however, they often incur substantial computational ov...
[ "Hongwei Chen", "Yishu Lei", "Dan Zhang", "Bo Ke", "Danxiang Zhu", "Xuyi Chen", "Yuxiang Lu", "Zhengjie Huang", "Shikun Feng", "Jingzhou He", "Yu Sun", "Hua Wu", "Haifeng Wang" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-10-11T00:00:00
https://arxiv.org/abs/2510.10293
https://arxiv.org/pdf/2510.10293v1
2510.10293
10.48550/arXiv.2510.10293
1
1
true
null
arXiv.org
0.5118
ff6d9005d733592d4c819f99aa54c50aa0a28d5242a0f3a23ddf9b2097ba09b5
[ "arxiv", "semantic_scholar" ]
Stop When Enough: Adaptive Early-Stopping for Chain-of-Thought Reasoning
Chain-of-Thought (CoT) reasoning has driven recent gains of large language models (LLMs) on reasoning-intensive tasks by externalizing intermediate steps. However, excessive or redundant reasoning -- so-called overthinking -- can increase inference costs and lead LLMs toward incorrect conclusions. In this paper, we pre...
[ "Renliang Sun", "Wei Cheng", "Dawei Li", "Haifeng Chen", "Wei Wang" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-11T00:00:00
https://arxiv.org/abs/2510.10103
https://arxiv.org/pdf/2510.10103v1
2510.10103
10.48550/arXiv.2510.10103
9
1
false
null
arXiv.org
0.3311
4ff9c55f61843a23110d4d8b3b6d492dcb9a866e332658def60b91ea98d1f269
[ "arxiv", "semantic_scholar" ]
Prompting Test-Time Scaling Is A Strong LLM Reasoning Data Augmentation
Large language models (LLMs) have demonstrated impressive reasoning capabilities when provided with chain-of-thought exemplars, but curating large reasoning datasets remains laborious and resource-intensive. In this work, we introduce Prompting Test-Time Scaling (P-TTS), a simple yet effective inference-time data augme...
[ "Sondos Mahmoud Bsharat", "Zhiqiang Shen" ]
[ "cs.CL", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-10-10T00:00:00
https://arxiv.org/abs/2510.09599
https://arxiv.org/pdf/2510.09599v1
2510.09599
10.48550/arXiv.2510.09599
2
0
true
https://github.com/VILA-Lab/PTTS
arXiv.org
0.51
67b74f94bb473a1f633a57fe03abaa8762e7742b81192eecce6ebd8b643babbb
[ "arxiv", "semantic_scholar" ]
Verifying Chain-of-Thought Reasoning via Its Computational Graph
Current Chain-of-Thought (CoT) verification methods predict reasoning correctness based on outputs (black-box) or activations (gray-box), but offer limited insight into why a computation fails. We introduce a white-box method: Circuit-based Reasoning Verification (CRV). We hypothesize that attribution graphs of correct...
[ "Zheng Zhao", "Yeskendir Koishekenov", "Xianjun Yang", "Naila Murray", "Nicola Cancedda" ]
[ "cs.CL", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-10-10T00:00:00
https://arxiv.org/abs/2510.09312
https://arxiv.org/pdf/2510.09312v2
2510.09312
10.48550/arXiv.2510.09312
17
3
false
null
arXiv.org
0.33
6d68fa5c0b3bf80e87c5957d0ed9712a6d7d97aa833740d55832e0d1b968abed
[ "arxiv", "semantic_scholar" ]
Parallel Test-Time Scaling for Latent Reasoning Models
Parallel test-time scaling (TTS) is a pivotal approach for enhancing large language models (LLMs), typically by sampling multiple token-based chains-of-thought in parallel and aggregating outcomes through voting or search. Recent advances in latent reasoning, where intermediate reasoning unfolds in continuous vector sp...
[ "Runyang You", "Yongqi Li", "Meng Liu", "Wenjie Wang", "Liqiang Nie", "Wenjie Li" ]
[ "cs.CL", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-10-09T00:00:00
https://arxiv.org/abs/2510.07745
https://arxiv.org/pdf/2510.07745v4
2510.07745
10.48550/arXiv.2510.07745
8
1
true
https://github.com/ModalityDance/LatentTTS
arXiv.org
0.5082
964ffb2f9b369104723c431b884722c9267fdc1e5fdd6fafc17cbb97291ff9df
[ "arxiv", "semantic_scholar" ]
Encode, Think, Decode: Scaling test-time reasoning with recursive latent thoughts
Most efforts to improve the reasoning capabilities of large language models (LLMs) involve either scaling the number of parameters and the size of training data, or scaling inference computation by letting models generate complex chains of thought. Motivated by interpretability studies showing that the crucial computat...
[ "Yeskendir Koishekenov", "Aldo Lipani", "Nicola Cancedda" ]
[ "cs.LG", "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2025-10-08T00:00:00
https://arxiv.org/abs/2510.07358
https://arxiv.org/pdf/2510.07358v1
2510.07358
10.48550/arXiv.2510.07358
15
1
false
null
arXiv.org
0.3277
09e562383364c44da5e4e408b90e5f9a8f178ed88753fba116d859d672376add
[ "arxiv", "semantic_scholar" ]
TS-Agent: Understanding and Reasoning Over Raw Time Series via Iterative Insight Gathering
Large language models (LLMs) exhibit strong symbolic and compositional reasoning, yet they struggle with time series question answering as the data is typically transformed into an LLM-compatible modality, e.g., serialized text, plotted images, or compressed time series embeddings. Such conversions impose representatio...
[ "Penghang Liu", "Elizabeth Fons", "Annita Vapsi", "Mohsen Ghassemi", "Svitlana Vyetrenko", "Daniel Borrajo", "Vamsi K. Potluru", "Manuela Veloso" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-10-08T00:00:00
https://arxiv.org/abs/2510.07432
https://arxiv.org/pdf/2510.07432v2
2510.07432
null
5
0
false
null
null
0.2085
ab392f0312bc097731e85582121fad7107fc078b8503fd2274eb553e8d99d339
[ "arxiv", "semantic_scholar" ]
ARISE: An Adaptive Resolution-Aware Metric for Test-Time Scaling Evaluation in Large Reasoning Models
Test-time scaling has emerged as a transformative paradigm for enhancing the performance of large reasoning models, enabling dynamic allocation of computational resources during inference. However, as the landscape of reasoning models rapidly expands, a critical question remains: how can we systematically compare and e...
[ "Zhangyue Yin", "Qiushi Sun", "Zhiyuan Zeng", "Zhiyuan Yu", "Qipeng Guo", "Xuanjing Huang", "Xipeng Qiu" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-10-07T00:00:00
https://arxiv.org/abs/2510.06014
https://arxiv.org/pdf/2510.06014v1
2510.06014
10.48550/arXiv.2510.06014
1
0
false
null
arXiv.org
0.3266
04bb0bb2ea623a194258974920cbb4f14fc46e69e16154c285cb27302b39e5e2
[ "arxiv", "semantic_scholar" ]
Test-Time Scaling of Reasoning Models for Machine Translation
Test-time scaling (TTS) has enhanced the performance of Reasoning Models (RMs) on various tasks such as math and coding, yet its efficacy in machine translation (MT) remains underexplored. This paper investigates whether increased inference-time computation improves translation quality. We evaluate 12 RMs across a dive...
[ "Zihao Li", "Shaoxiong Ji", "Jörg Tiedemann" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-07T00:00:00
https://arxiv.org/abs/2510.06471
https://arxiv.org/pdf/2510.06471v2
2510.06471
10.48550/arXiv.2510.06471
3
0
false
null
Conference of the European Chapter of the Association for Computational Linguistics
0.3266
ab937065e6d5a3521849961af1e6065cdf7a1ff385f54775f1a4dd21d97c56c0
[ "arxiv", "semantic_scholar" ]
The Valley of Code Reasoning: Scaling Knowledge Distillation of Large Language Models
Distilling the thinking traces of a Large Language Model (LLM) with reasoning capabilities into a smaller model has been proven effective. Yet, there is a scarcity of work done on how model performances scale with the quantity of distillation data. In this work, we study the scaling trend of distilling competitive codi...
[ "Muyu He", "Muhammad Ali Shafique", "Anand Kumar", "Tsach Mackey", "Nazneen Rajani" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-10-07T00:00:00
https://arxiv.org/abs/2510.06101
https://arxiv.org/pdf/2510.06101v1
2510.06101
10.48550/arXiv.2510.06101
0
0
false
null
arXiv.org
0.3266
0efedbcd75c1655e621e130fbc4f660f777581155bddafbcec00a3cbd3637348
[ "arxiv", "semantic_scholar" ]
VChain: Chain-of-Visual-Thought for Reasoning in Video Generation
Recent video generation models can produce smooth and visually appealing clips, but they often struggle to synthesize complex dynamics with a coherent chain of consequences. Accurately modeling visual outcomes and state transitions over time remains a core challenge. In contrast, large language and multimodal models (e...
[ "Ziqi Huang", "Ning Yu", "Gordon Chen", "Haonan Qiu", "Paul Debevec", "Ziwei Liu" ]
[ "cs.CV" ]
[ "Computer Science" ]
2025-10-06T00:00:00
https://arxiv.org/abs/2510.05094
https://arxiv.org/pdf/2510.05094v2
2510.05094
10.48550/arXiv.2510.05094
22
0
false
null
arXiv.org
0.3404
550572696168a1b7cac328d9a248dfc79d18ca7f7cf47b8814551e02639d4782
[ "arxiv", "semantic_scholar" ]
Thinking on the Fly: Test-Time Reasoning Enhancement via Latent Thought Policy Optimization
Recent advancements in Large Language Models (LLMs) have shifted from explicit Chain-of-Thought (CoT) reasoning to more efficient latent reasoning, where intermediate thoughts are represented as vectors rather than text. However, latent reasoning can be brittle on challenging, out-of-distribution tasks where robust rea...
[ "Wengao Ye", "Yan Liang", "Lianlei Shan" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-10-05T00:00:00
https://arxiv.org/abs/2510.04182
https://arxiv.org/pdf/2510.04182v4
2510.04182
10.48550/arXiv.2510.04182
6
1
false
null
arXiv.org
0.3243
0f995e04e5a4aa896ccfd961cb22c128fbaa2790a578065c10f3d66c98976dc6
[ "arxiv", "semantic_scholar" ]
Scaling Code-Assisted Chain-of-Thoughts and Instructions for Model Reasoning
Reasoning capability is pivotal for Large Language Models (LLMs) to solve complex tasks, yet achieving reliable and scalable reasoning remains challenging. While Chain-of-Thought (CoT) prompting has become a mainstream approach, existing methods often suffer from uncontrolled generation, insufficient quality, and limit...
[ "Honglin Lin", "Qizhi Pei", "Xin Gao", "Zhuoshi Pan", "Yu Li", "Juntao Li", "Conghui He", "Lijun Wu" ]
[ "cs.CL", "cs.PL" ]
[ "Computer Science" ]
2025-10-05T00:00:00
https://arxiv.org/abs/2510.04081
https://arxiv.org/pdf/2510.04081v1
2510.04081
10.48550/arXiv.2510.04081
10
1
false
null
arXiv.org
0.3243
90732757a55f26c6464d272da90b41c4e8dd1195f9dcb9a1ef513a55c4559668
[ "arxiv", "semantic_scholar" ]
Understanding the Role of Training Data in Test-Time Scaling
Test-time scaling improves the reasoning capabilities of large language models (LLMs) by allocating extra compute to generate longer Chains-of-Thoughts (CoTs). This enables models to tackle more complex problem by breaking them down into additional steps, backtracking, and correcting mistakes. Despite its strong perfor...
[ "Adel Javanmard", "Baharan Mirzasoleiman", "Vahab Mirrokni" ]
[ "cs.AI", "cs.LG", "stat.ML" ]
[ "Computer Science", "Mathematics" ]
2025-10-04T00:00:00
https://arxiv.org/abs/2510.03605
https://arxiv.org/pdf/2510.03605v2
2510.03605
10.48550/arXiv.2510.03605
4
1
false
null
arXiv.org
0.3231
0c0e005ea5a1b8cefa96413a2a1735da5aeaa46c0092e7b3e38b4bbac298356b
[ "arxiv", "semantic_scholar" ]
RoiRL: Efficient, Self-Supervised Reasoning with Offline Iterative Reinforcement Learning
Reinforcement learning (RL) is central to improving reasoning in large language models (LLMs) but typically requires ground-truth rewards. Test-Time Reinforcement Learning (TTRL) removes this need by using majority-vote rewards, but relies on heavy online RL and incurs substantial computational cost. We propose RoiRL: ...
[ "Aleksei Arzhantsev", "Otmane Sakhi", "Flavian Vasile" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-10-03T00:00:00
https://arxiv.org/abs/2510.02892
https://arxiv.org/pdf/2510.02892v1
2510.02892
10.48550/arXiv.2510.02892
1
0
false
null
arXiv.org
0.322
83a245811c6cfec335c78e82cb17472128ef6b91ce6a4231303e1ade93a2128d
[ "arxiv", "semantic_scholar" ]
Eliciting Chain-of-Thought Reasoning for Time Series Analysis using Reinforcement Learning
Complex numerical time series analysis often demands multi-step reasoning capabilities beyond current models' reach. Tasks like medical diagnosis and weather forecasting require sequential reasoning processes - including counterfactual analysis, logical deduction, knowledge application, and multi-modal contextual integ...
[ "Felix Parker", "Nimeesha Chan", "Chi Zhang", "Kimia Ghobadi" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-10-01T00:00:00
https://arxiv.org/abs/2510.01116
https://arxiv.org/pdf/2510.01116v2
2510.01116
10.48550/arXiv.2510.01116
5
0
false
null
arXiv.org
0.3197
5842002303bbb250f6acd4286b76df6e5a57d20d0be2ff6203afcacf98c2d269
[ "arxiv", "semantic_scholar" ]
Typed Chain-of-Thought: A Curry-Howard Framework for Verifying LLM Reasoning
While Chain-of-Thought (CoT) prompting enhances the reasoning capabilities of large language models, the faithfulness of the generated rationales remains an open problem for model interpretability. We propose a novel theoretical lens for this problem grounded in the Curry-Howard correspondence, which posits a direct re...
[ "Elija Perrier" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-10-01T00:00:00
https://arxiv.org/abs/2510.01069
https://arxiv.org/pdf/2510.01069v1
2510.01069
10.48550/arXiv.2510.01069
3
0
false
null
arXiv.org
0.3197
2ca47670badc8501fb5f6092359cd95a1f59e0922fd2289021f1b4736107b428
[ "arxiv", "semantic_scholar" ]
ThinKV: Thought-Adaptive KV Cache Compression for Efficient Reasoning Models
The long-output context generation of large reasoning models enables extended chain of thought (CoT) but also drives rapid growth of the key-value (KV) cache, quickly overwhelming GPU memory. To address this challenge, we propose ThinKV, a thought-adaptive KV cache compression framework. ThinKV is based on the observat...
[ "Akshat Ramachandran", "Marina Neseem", "Charbel Sakr", "Rangharajan Venkatesan", "Brucek Khailany", "Tushar Krishna" ]
[ "cs.LG" ]
[ "Computer Science" ]
2025-10-01T00:00:00
https://arxiv.org/abs/2510.01290
https://arxiv.org/pdf/2510.01290v2
2510.01290
10.48550/arXiv.2510.01290
8
1
false
null
arXiv.org
0.3197
d2686c39aeb3a412985d635072ac3e58db65de49c9dc391d3e2385e8461848f0
[ "arxiv", "semantic_scholar" ]
ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory
With the growing adoption of large language model agents in persistent real-world roles, they naturally encounter continuous streams of tasks. A key limitation, however, is their failure to learn from the accumulated interaction history, forcing them to discard valuable insights and repeat past errors. We propose Reaso...
[ "Siru Ouyang", "Jun Yan", "I-Hung Hsu", "Yanfei Chen", "Ke Jiang", "Zifeng Wang", "Rujun Han", "Long T. Le", "Samira Daruki", "Xiangru Tang", "Vishy Tirumalashetty", "George Lee", "Mahsan Rofouei", "Hangfei Lin", "Jiawei Han", "Chen-Yu Lee", "Tomas Pfister" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2025-09-29T00:00:00
https://arxiv.org/abs/2509.25140
https://arxiv.org/pdf/2509.25140v2
2509.25140
10.48550/arXiv.2509.25140
126
26
true
https://github.com/google-research/reasoning-bank
arXiv.org
0.7157
bbc52395a40f3dc504129637afd9765ff6577fd4e67c0b5006d88af7a6824a5d
[ "arxiv", "semantic_scholar" ]
ARS: Adaptive Reasoning Suppression for Efficient Large Reasoning Language Models
Large Reasoning Language Models (LRLMs or LRMs) demonstrate remarkable capabilities in complex reasoning tasks, but suffer from significant computational inefficiencies due to overthinking phenomena. Existing efficient reasoning methods face the challenge of balancing reasoning quality with inference cost reduction. We...
[ "Dongqi Zheng" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2025-09-29T00:00:00
https://arxiv.org/abs/2510.00071
https://arxiv.org/pdf/2510.00071v2
2510.00071
10.48550/arXiv.2510.00071
0
0
false
null
arXiv.org
0.3174
afff02ab49b0d8ba366fa5f5109780dfd683d163e2e12102ec145e198523840e
[ "arxiv", "semantic_scholar" ]
AdvChain: Adversarial Chain-of-Thought Tuning for Robust Safety Alignment of Large Reasoning Models
Large Reasoning Models (LRMs) have demonstrated remarkable capabilities in complex problem-solving through Chain-of-Thought (CoT) reasoning. However, the multi-step nature of CoT introduces new safety challenges that extend beyond conventional language model alignment. We identify a failure mode in current safety CoT t...
[ "Zihao Zhu", "Xinyu Wu", "Gehan Hu", "Siwei Lyu", "Ke Xu", "Baoyuan Wu" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2025-09-29T00:00:00
https://arxiv.org/abs/2509.24269
https://arxiv.org/pdf/2509.24269v1
2509.24269
10.48550/arXiv.2509.24269
5
0
false
null
arXiv.org
0.3174
5782b7c8f7947d8de5d712db30256b7dd116721658f4508c7c687231787f7b20
[ "arxiv", "semantic_scholar" ]
Scaling LLM Test-Time Compute with Mobile NPU on Smartphones
Deploying Large Language Models (LLMs) on mobile devices faces the challenge of insufficient performance in smaller models and excessive resource consumption in larger ones. This paper highlights that mobile Neural Processing Units (NPUs) have underutilized computational resources, particularly their matrix multiplicat...
[ "Zixu Hao", "Jianyu Wei", "Tuowei Wang", "Minxing Huang", "Huiqiang Jiang", "Shiqi Jiang", "Ting Cao", "Ju Ren" ]
[ "cs.DC", "cs.AI" ]
[ "Computer Science" ]
2025-09-27T00:00:00
https://arxiv.org/abs/2509.23324
https://arxiv.org/pdf/2509.23324v1
2509.23324
10.48550/arXiv.2509.23324
6
1
false
null
European Conference on Computer Systems
0.3151
d78efc000243b8495561c47a974f12aa2c52886e4127b18a03d58143f64be7d7
[ "arxiv", "semantic_scholar" ]
Retrieval-of-Thought: Efficient Reasoning via Reusing Thoughts
Large reasoning models improve accuracy by producing long reasoning traces, but this inflates latency and cost, motivating inference-time efficiency. We propose Retrieval-of-Thought (RoT), which reuses prior reasoning as composable ``thought" steps to guide new problems. RoT organizes steps into a thought graph with se...
[ "Ammar Ahmed", "Azal Ahmad Khan", "Ayaan Ahmad", "Sheng Di", "Zirui Liu", "Ali Anwar" ]
[ "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-09-26T00:00:00
https://arxiv.org/abs/2509.21743
https://arxiv.org/pdf/2509.21743v2
2509.21743
10.48550/arXiv.2509.21743
7
0
false
null
arXiv.org
0.314
a038be76a893bbf1f10aa17dcc1786218ec34fdee0677d6e4199f6ffab08b7c8
[ "arxiv", "semantic_scholar" ]
Dual-Head Reasoning Distillation: Improving Classifier Accuracy with Train-Time-Only Reasoning
Chain-of-Thought (CoT) prompting often improves classification accuracy, but it introduces a significant throughput penalty with rationale generation (Wei et al., 2022; Cheng and Van Durme, 2024). To resolve this trade-off, we introduce Dual-Head Reasoning Distillation (DHRD), a simple training method for decoder-only ...
[ "Jillian Xu", "Dylan Zhou", "Vinay Shukla", "Yang Yang", "Junrui Ruan", "Shuhuai Lin", "Wenfei Zou", "Yinxiao Liu", "Karthik Lakshmanan" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-09-25T00:00:00
https://arxiv.org/abs/2509.21487
https://arxiv.org/pdf/2509.21487v2
2509.21487
10.48550/arXiv.2509.21487
3
0
false
null
arXiv.org
0.3128
b1a06449b3d57105e9692591ff6e78ef29b89fa83fe3efdbc4bf906a377ed051
[ "arxiv", "semantic_scholar" ]
A Formal Comparison Between Chain of Thought and Latent Thought
Chain of thought (CoT) elicits reasoning in large language models by explicitly generating intermediate tokens. In contrast, latent thought reasoning operates directly in the continuous latent space, enabling computation beyond discrete linguistic representations. While both approaches exploit iterative computation, th...
[ "Kevin Xu", "Issei Sato" ]
[ "cs.AI", "cs.CL", "cs.LG" ]
[ "Computer Science" ]
2025-09-25T00:00:00
https://arxiv.org/abs/2509.25239
https://arxiv.org/pdf/2509.25239v3
2509.25239
10.48550/arXiv.2509.25239
6
0
false
null
arXiv.org
0.3128
00c9b68c0160766401e6a7b0239aa011439a33fc66075627a4ca322aae7ffe35
[ "arxiv", "semantic_scholar" ]
Expanding Reasoning Potential in Foundation Model by Learning Diverse Chains of Thought Patterns
Recent progress in large reasoning models for challenging mathematical reasoning has been driven by reinforcement learning (RL). Incorporating long chain-of-thought (CoT) data during mid-training has also been shown to substantially improve reasoning depth. However, current approaches often utilize CoT data indiscrimin...
[ "Xuemiao Zhang", "Can Ren", "Chengying Tu", "Rongxiang Weng", "Shuo Wang", "Hongfei Yan", "Jingang Wang", "Xunliang Cai" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2025-09-25T00:00:00
https://arxiv.org/abs/2509.21124
https://arxiv.org/pdf/2509.21124v3
2509.21124
10.48550/arXiv.2509.21124
3
0
false
null
arXiv.org
0.3128
6c7498d45142e3780a3f1cb20c68ada3676df4d46f7692f2470bba43e2443693
[ "arxiv", "semantic_scholar" ]
Mitigating Strategy-Selection Bias in Reasoning for More Effective Test-Time Scaling
Test-time scaling (TTS) has been shown to improve the performance of large language models (LLMs) by sampling and aggregating diverse reasoning paths. However, existing research has overlooked a critical issue: selection bias of reasoning strategies during scaling. Specifically, when generating reasoning processes, LLM...
[ "Zongqian Wu", "Baoduo Xu", "Tianyu Li", "Zhu Sun", "Xiaofeng Zhu", "Lei Feng" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-09-22T00:00:00
https://arxiv.org/abs/2509.17905
https://arxiv.org/pdf/2509.17905v2
2509.17905
10.48550/arXiv.2509.17905
2
0
false
null
arXiv.org
0.3094
8aca79dbbace566584c40b711457649dd1d5cb69acc57bf3f62582da44e06c66
[ "arxiv", "semantic_scholar" ]
Conversational Orientation Reasoning: Egocentric-to-Allocentric Navigation with Multimodal Chain-of-Thought
Conversational agents must translate egocentric utterances (e.g., "on my right") into allocentric orientations (N/E/S/W). This challenge is particularly critical in indoor or complex facilities where GPS signals are weak and detailed maps are unavailable. While chain-of-thought (CoT) prompting has advanced reasoning in...
[ "Yu Ti Huang" ]
[ "cs.LG", "cs.AI", "cs.CL", "cs.RO" ]
[ "Computer Science" ]
2025-09-20T00:00:00
https://arxiv.org/abs/2509.18200
https://arxiv.org/pdf/2509.18200v1
2509.18200
10.48550/arXiv.2509.18200
0
0
false
null
arXiv.org
0.3071
64a51bd6d2d8089307a8253f94dcc6b536f2f75ffd68fb17484a4c1567b1f9e9
[ "arxiv", "semantic_scholar" ]
Slim-SC: Thought Pruning for Efficient Scaling with Self-Consistency
Recently, Test-Time Scaling (TTS) has gained increasing attention for improving LLM reasoning performance at test time without retraining the model. A notable TTS technique is Self-Consistency (SC), which generates multiple reasoning chains in parallel and selects the final answer via majority voting. While effective, ...
[ "Colin Hong", "Xu Guo", "Anand Chaanan Singh", "Esha Choukse", "Dmitrii Ustiugov" ]
[ "cs.CL", "cs.AI", "cs.LG" ]
[ "Computer Science" ]
2025-09-17T00:00:00
https://arxiv.org/abs/2509.13990
https://arxiv.org/pdf/2509.13990v1
2509.13990
10.48550/arXiv.2509.13990
12
5
false
null
Conference on Empirical Methods in Natural Language Processing
0.3891
99a0015b0a7d4c9436d99bbeeac59213b8b563ea2e6da3536d80cd6c9a01328d
[ "arxiv", "semantic_scholar" ]
LTA-thinker: Latent Thought-Augmented Training Framework for Large Language Models on Complex Reasoning
Complex Reasoning in Large Language Models can be dynamically optimized using Test-Time Scaling (TTS) to mitigate Overthinking. Methods such as Coconut, SoftCoT and its variant are effective in continuous latent space inference, the core bottleneck still lies in the efficient generation and utilization of high-quality ...
[ "Jiaqi Wang", "Binquan Ji", "Haibo Luo", "Yiyang Qi", "Ruiting Li", "Huiyan Wang", "Yuantao Han", "Cangyi Yang", "jiaxu Zhang", "Feiliang Ren" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-09-16T00:00:00
https://arxiv.org/abs/2509.12875
https://arxiv.org/pdf/2509.12875v3
2509.12875
10.48550/arXiv.2509.12875
4
1
false
null
arXiv.org
0.3025
ca4eb42dea342143ca5eef4ebb2047a039102531dc9348dae72785a8be4d20e5
[ "arxiv", "semantic_scholar" ]
Reasoning Models Can be Accurately Pruned Via Chain-of-Thought Reconstruction
Reasoning language models such as DeepSeek-R1 produce long chain-of-thought traces during inference time which make them costly to deploy at scale. We show that using compression techniques such as neural network pruning produces greater performance loss than in typical language modeling tasks, and in some cases can ma...
[ "Ryan Lucas", "Kayhan Behdin", "Zhipeng Wang", "Qingquan Song", "Shao Tang", "Rahul Mazumder" ]
[ "cs.AI" ]
[ "Computer Science" ]
2025-09-15T00:00:00
https://arxiv.org/abs/2509.12464
https://arxiv.org/pdf/2509.12464v2
2509.12464
10.48550/arXiv.2509.12464
1
0
true
https://github.com/RyanLucas3/RAC
arXiv.org
0.4657
2af4a7e334a41eb6e5b3c79a333e3d337ba24e39da84413d54f3228899acfd64
[ "arxiv", "semantic_scholar" ]
K2-Think: A Parameter-Efficient Reasoning System
K2-Think is a reasoning system that achieves state-of-the-art performance with a 32B parameter model, matching or surpassing much larger models like GPT-OSS 120B and DeepSeek v3.1. Built on the Qwen2.5 base model, our system shows that smaller models can compete at the highest levels by combining advanced post-training...
[ "Zhoujun Cheng", "Richard Fan", "Shibo Hao", "Taylor W. Killian", "Haonan Li", "Suqi Sun", "Hector Ren", "Alexander Moreno", "Daqian Zhang", "Tianjun Zhong", "Yuxin Xiong", "Yuanzhe Hu", "Yutao Xie", "Xudong Han", "Yuqi Wang", "Varad Pimpalkhute", "Yonghao Zhuang", "Aaryamonvikram ...
[ "cs.LG" ]
[ "Computer Science" ]
2025-09-09T00:00:00
https://arxiv.org/abs/2509.07604
https://arxiv.org/pdf/2509.07604v3
2509.07604
10.48550/arXiv.2509.07604
8
0
true
null
arXiv.org
0.4551
0492f7dbb7295ba2b9666a96d705d72ac11bdb9dba3e42650f253687a2baccb1
[ "arxiv", "semantic_scholar" ]
Test-Time Scaling in Reasoning Models Is Not Effective for Knowledge-Intensive Tasks Yet
Test-time scaling increases inference-time computation by allowing models to generate long reasoning chains, and has improved performance across many domains. However, in this work, we show that this approach is not yet effective for knowledge-intensive tasks. We evaluate 14 reasoning models on two knowledge-intensive ...
[ "James Xu Zhao", "Bryan Hooi", "See-Kiong Ng" ]
[ "cs.AI", "cs.CL", "cs.LG" ]
[ "Computer Science" ]
2025-09-08T00:00:00
https://arxiv.org/abs/2509.06861
https://arxiv.org/pdf/2509.06861v2
2509.06861
10.48550/arXiv.2509.06861
10
0
true
https://github.com/XuZhao0/tts-knowledge
arXiv.org
0.4533
30a81d504c8af4c2229e4982d7e971d2c8ca9c567cecdf2ecb292af21d961fa4
[ "arxiv", "semantic_scholar" ]
Less is More Tokens: Efficient Math Reasoning via Difficulty-Aware Chain-of-Thought Distillation
Chain-of-thought reasoning, while powerful, can produce unnecessarily verbose output for simpler problems. We present a framework for difficulty-aware reasoning that teaches models to dynamically adjust reasoning depth based on problem complexity. Remarkably, we show that models can be endowed with such dynamic inferen...
[ "Abdul Waheed", "Chancharik Mitra", "Laurie Z. Wang", "Deva Ramanan", "Bhiksha Raj" ]
[ "cs.CL" ]
[ "Computer Science" ]
2025-09-05T00:00:00
https://arxiv.org/abs/2509.05226
https://arxiv.org/pdf/2509.05226v1
2509.05226
10.48550/arXiv.2509.05226
0
0
false
null
arXiv.org
0.2899
29d52e9bc0c8b1dea677661900835320f195e3c6de789226a6c0cd6c948b183f
[ "arxiv", "semantic_scholar" ]
Chain or tree? Re-evaluating complex reasoning from the perspective of a matrix of thought
Large Language Models (LLMs) face significant accuracy degradation due to insufficient reasoning ability when dealing with complex and abstract tasks. Thought structures such as Chain of Thought (CoT) and Tree of Thought (ToT) focus on enhancing the reasoning capability of LLMs. However, they suffer from inherent drawb...
[ "Fengxiao Tang", "Yufeng Li", "Zongzong Wu", "Ming Zhao" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-09-04T00:00:00
https://arxiv.org/abs/2509.03918
https://arxiv.org/pdf/2509.03918v2
2509.03918
10.48550/arXiv.2509.03918
1
0
true
https://github.com/lyfiter/mtqa
arXiv.org
0.4462
4c97fb75a345424b91c894df5551ab0f00d354a9dc99f1988028ba5873d78081
[ "arxiv", "semantic_scholar" ]
Why Does Reasoning Length Converge? Unveiling the Underfitting-Overfitting Trade-off in Chain-of-Thought
Test-time scaling, primarily manifested through multi-step Chain-of-Thought (CoT) reasoning via Reinforcement Learning (RL), has emerged as a pivotal paradigm for enhancing the reasoning capabilities of Large Language Models (LLMs). However, a significant theoretical gap persists: traditional token-level analysis fails...
[ "Zeyu Gan", "Hao Yi", "Yong Liu" ]
[ "cs.AI", "cs.CL" ]
[ "Computer Science" ]
2025-09-04T00:00:00
https://arxiv.org/abs/2509.04027
https://arxiv.org/pdf/2509.04027v4
2509.04027
null
2
1
false
null
null
0.1837
e013a8d45247186efb8662c143dabd60f1efc453be131dc361e864180b3af797
[ "arxiv", "semantic_scholar" ]
ParaThinker: Native Parallel Thinking as a New Paradigm to Scale LLM Test-time Compute
Recent advances in Large Language Models (LLMs) have been driven by test-time compute scaling - a strategy that improves reasoning by generating longer, sequential thought processes. While effective, this approach encounters a significant bottleneck as computation increases, where further computation offers only margin...
[ "Hao Wen", "Yifan Su", "Feifei Zhang", "Yunxin Liu", "Yunhao Liu", "Ya-Qin Zhang", "Yuanchun Li" ]
[ "cs.CL", "cs.AI" ]
[ "Computer Science" ]
2025-08-30T00:00:00
https://arxiv.org/abs/2509.04475
https://arxiv.org/pdf/2509.04475v1
2509.04475
10.48550/arXiv.2509.04475
34
1
false
null
arXiv.org
0.386
b686dc1fffbef85733df570f6eb505850111cfdf23d02217f88e789114384153
[ "arxiv", "semantic_scholar" ]
SQL-of-Thought: Multi-agentic Text-to-SQL with Guided Error Correction
Converting natural language queries into SQL queries is a crucial challenge in both industry and academia, aiming to increase access to databases and large-scale applications. This work examines how in-context learning and chain-of-thought can be utilized to develop a robust solution for text-to-SQL systems. We propose...
[ "Saumya Chaturvedi", "Aman Chadha", "Laurent Bindschaedler" ]
[ "cs.DB", "cs.LG" ]
[ "Computer Science" ]
2025-08-30T00:00:00
https://arxiv.org/abs/2509.00581
https://arxiv.org/pdf/2509.00581v2
2509.00581
10.48550/arXiv.2509.00581
8
0
true
https://github.com/shollercoaster/SQL-of-Thought
arXiv.org
0.4374