List your company

Faithful Calibration on PQA

89cMFG*Qwen3-1.7B + RLMF + Rewr.
45.3256.666879.34Jun 30, 2026

Evaluation results

MethodLinks
Qwen3-1.7B + RLMF + Rewr.Variant=RLMF + Rewriting2026.0689
Llama3.1-8B-Ins + RLMFVariant=RLMF2026.0685
Qwen3-8B + RLMFVariant=RLMF2026.0685
Qwen3-1.7B + RLMFVariant=RLMF2026.0683
Llama3.1-8B-Ins + RLVariant=RL2026.0682
Qwen3-8B + RLMF + Rewr.Variant=RLMF + Rewriting2026.0682
Llama3.1-8B-Ins + RLMF + Rewr.Variant=RLMF + Rewriting2026.0681
Qwen3-4B + RLMFVariant=RLMF2026.0681
Qwen3-4B + RLVariant=RL2026.0680
Qwen3-4B + RLMF + Rewr.Variant=RLMF + Rewriting2026.0680
Qwen3-1.7B + RLVariant=RL2026.0678
Qwen3-8B + RLVariant=RL2026.0675
Llama3.1-8B-Ins + FUTVariant=FUT2026.0669
Llama3.1-8B-Ins + MetaFaithVariant=MetaFaith2026.0668
Gemini-3.1-ProVariant=Base2026.0662
Llama3.1-8B-InsVariant=Base2026.0660
Gemini-3-FlashVariant=Base2026.0659
Qwen3-8B + FUTVariant=FUT2026.0657
Qwen3-1.7BVariant=Base2026.0657
Qwen3-1.7B + FUTVariant=FUT2026.0655
Qwen3-8BVariant=Base2026.0653
Qwen3-8B + MetaFaithVariant=MetaFaith2026.0653
Qwen3-4BVariant=Base2026.0653
Qwen3-1.7B + MetaFaithVariant=MetaFaith2026.0650
GPT-5Variant=Base2026.0650
Qwen3-4B + MetaFaithVariant=MetaFaith2026.0648
Qwen3-4B + FUTVariant=FUT2026.0647