ResearchBenchmarksMulti-hop Fact Verification on HoVer 4-HopFollow63Macro-F1MERMAID51.5654.5357.560.47Jan 29, 2026Evaluation ResultsMethodMethodLinksMacro-F1MERMAIDLLM=GPT-4oLLM=GPT-4o2026.0163MERMAIDLLM=GPT-5 MiniLLM=GPT-5 Mini2026.0162FOLK2026.0160MERMAIDLLM=OSS-120BLLM=OSS-120B2026.0160MERMAIDLLM=Qwen-2.5-70BLLM=Qwen-2.5-70B2026.0157ProgramFC2026.0153Self-Ask2026.0152