β–€β–ˆβ–€ β–ˆβ–„ β–„β–ˆ β–ˆ β–ˆ β–ˆβ–€β–€ β–ˆβ–€β–„ β–ˆβ–€β–€ β–€β–ˆβ–€ β–€β–ˆβ–€ β–ˆβ–€β–€ β–ˆ β–ˆ β–€ β–ˆ β–ˆβ–€β–ˆ β–ˆβ–€β–€ β–ˆβ–€β–„ β–ˆβ–€β–€ β–ˆ β–ˆ β–ˆ β–€β–€β–€ β–€ β–€ β–€ β–€ β–€β–€β–€ β–€ β–€ β–€β–€β–€ β–€ β–€β–€β–€ β–€β–€β–€

Abliterated/Heretic KingNish/Qwen2.5-0.5b-RBase

Check Quants

Refusals (this model): 4/100
Original (KingNish/Qwen2.5-0.5b-RBase): 27/100
KL divergence: 0.0292

Parameters
direction_index = 18.25
attn.o_proj.max_weight = 1.08
attn.o_proj.max_weight_position = 22.92
attn.o_proj.min_weight = 0.37
attn.o_proj.min_weight_distance = 13.76
mlp.down_proj.max_weight = 1.44
mlp.down_proj.max_weight_position = 15.22
mlp.down_proj.min_weight = 0.84
mlp.down_proj.min_weight_distance = 5.31


Downloads last month
13
Safetensors
Model size
0.5B params
Tensor type
F16
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for hereticness/Heretic-Qwen2.5-0.5b-RBase

Base model

Qwen/Qwen2.5-0.5B
Finetuned
(1)
this model
Quantizations
2 models