List your company

Reasoning on GPQA-Diamond (generalization)

41.6AccuracyAutoTTS
40.97641.13841.341.462May 8, 2026

Evaluation results

MethodLinks
AutoTTSType=Discovered, Base Model=Qwen3-1.7B, β=12026.0541.6270,100
AutoTTSType=Discovered, Base Model=Qwen3-1.7B, β=0.52026.0541.6151,000
SC@64Type=Handcrafted, Base Model=Qwen3-1.7B2026.0541.3510,000
ESCType=Handcrafted, Base Model=Qwen3-1.7B2026.0541.3391,600
ASCType=Handcrafted, Base Model=Qwen3-1.7B2026.0541186,300