List your company

Claim Verification on SciTab-OD

77Macro F1Llama-400B
35.446.25767.8Apr 19, 2026

Evaluation results

MethodLinks
Llama-400BReasoning=w COT2026.0477
MACEBackbone=Qw-235B, Configuration=Pm + Em + Vm2026.0476
Llama-70BReasoning=w COT2026.0474
Qwen-235BReasoning=w COT2026.0474
DeepSeek-V3.1Reasoning=w COT2026.0474
DeepSeek-70BReasoning=wo COT2026.0473
Llama-400BReasoning=wo COT2026.0473
DeepSeek-70BReasoning=w COT2026.0472
Qwen-72BReasoning=w COT2026.0472
GPT-4oReasoning=w COT2026.0471
Claude-3.5-SonnetReasoning=w COT2026.0471
Claude-3.5-SonnetReasoning=wo COT2026.0471
MACEBackbone=Qw-72B, Configuration=Pm + Em + Vm2026.0471
Qwen-72BReasoning=wo COT2026.0469
Gemma-12BReasoning=w COT2026.0468
Qwen-235BReasoning=wo COT2026.0468
MACEBackbone=Ll-8B, Configuration=Pm + Em + Vm'2026.0466
Mistral-7BReasoning=w COT2026.0465
Llama-70BReasoning=wo COT2026.0465
Gemma-12BReasoning=wo COT2026.0464
GPT-4oReasoning=wo COT2026.0460
MACEBackbone=Mt-7B, Configuration=Pm + Em + Vm'2026.0460
Llama-8BReasoning=wo COT2026.0459
Mistral-7BReasoning=wo COT2026.0458
Llama-3BReasoning=w COT2026.0457
Llama-8BReasoning=w COT2026.0456
DeepSeek-V3.1Reasoning=wo COT2026.0456
Llama-3BReasoning=wo COT2026.0437