List your company

Mathematical Reasoning on AIME (Acceptance length τ)

4.757Acceptance Length (τ)Llama-3-8B
1.920922.657213.39354.12979Jun 10, 2026

Evaluation results

MethodLinks
Llama-3-8Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=402026.064.757
Llama 3.2-3Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=402026.064.197
Llama-3-8Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=02026.063.465
Llama 3.2-3Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=02026.063.455
Llama 3.2-3Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.063.356
Llama-3-8Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.063.152
Llama 3.2-3Bγ (loss-decay constant)=None, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.063.003
Llama-3-8Bγ (loss-decay constant)=None, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.062.705
Qwen3-8Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=402026.062.558
Qwen3-4Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=402026.062.554
Qwen3-4Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=02026.062.357
Qwen3-8Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0.3, αc (chain-loss coefficient)=02026.062.351
Qwen3-4Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.062.253
Qwen3-8Bγ (loss-decay constant)=10, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.062.233
Qwen3-4Bγ (loss-decay constant)=None, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.062.133
Qwen3-8Bγ (loss-decay constant)=None, αf (first-error focal coefficient)=0, αc (chain-loss coefficient)=02026.062.03