ResearchBenchmarksOutcome Reasoning on CRASSFollow92.1M' (F1 Mean)GPT-569.2275.1681.187.04May 17, 2025Evaluation ResultsMethodMethodLinksM' (F1 Mean)Y' (F1 Mean)GPT-5Model=GPT-5Model=GPT-52025.0592.188GPT-o4Model=GPT-o4Model=GPT-o42025.0590.586.2Llama4-MModel=Llama4-MModel=Llama4-M2025.0584.979.5DeepSeekModel=DeepSeekModel=DeepSeek2025.0582.977.1Gemini2.5Model=Gemini2.5Model=Gemini2.52025.0581.775.2Qwen3Model=Qwen3Model=Qwen32025.0580.573.9Llama4-SModel=Llama4-SModel=Llama4-S2025.0570.163.5