ResearchBenchmarksMulti-hop Fact Verification on HoVer 2-HopFollow71Macro F1ProgramFC53.3257.9162.567.09Jan 29, 2026Evaluation ResultsMethodMethodLinksMacro F1ProgramFC2026.0171MERMAIDLLM=GPT-4oLLM=GPT-4o2026.0171MERMAIDLLM=GPT-5 MiniLLM=GPT-5 Mini2026.0170MERMAIDLLM=OSS-120BLLM=OSS-120B2026.0168FOLK2026.0166MERMAIDLLM=Qwen-2.5-70BLLM=Qwen-2.5-70B2026.0165Self-Ask2026.0154