List your company

Mathematical Reasoning on AMC23 (pass@1 and pass@16)

76.4Pass@1ExTra (full)
71.7272.93574.1575.365Jun 23, 2026

Evaluation results

MethodLinks
ExTra (full)Backbone=Qwen3-1.7B2026.0676.495
ExTra-NBackbone=Qwen3-1.7B, Variant=Correctness-gated novelty reward only2026.0674.797.5
DAPOBackbone=Qwen3-1.7B2026.0672.592.5
ExTra-RBackbone=Qwen3-1.7B, Variant=Entropy-guided prefix regeneration only2026.0672.595
GRPOBackbone=Qwen3-1.7B2026.0671.992.5