List your company

STEM Theorem Question Answering on TheoremQA

4.4Acceptance LengthTTS
1.0721.9362.83.664May 10, 2026

Evaluation results

MethodLinks
TTSTarget Model=Qwen/Qwen3-4B2026.054.423.4
TTSTarget Model=Qwen/Qwen3-8B2026.054.322.8
TTSTarget Model=Qwen/Qwen3.5-122B2026.054.225.1
DFlashTarget Model=Qwen/Qwen3-4B2026.053.6—
DFlashTarget Model=Qwen/Qwen3-8B2026.053.5—
DFlashTarget Model=Qwen/Qwen3.5-122B2026.053.3—
TTSModel=Llama3.1-8B2026.051.956.9
TTSModel=Qwen/Qwen3-8B2026.051.916.8
TTSModel=Qwen/Qwen3-32B2026.051.821.9
EAGLE-3Model=Qwen/Qwen3-8B2026.051.6—
EAGLE-3Model=Qwen/Qwen3-32B2026.051.5—
EAGLE-3Model=Llama3.1-8B2026.051.2—