ResearchBenchmarksData Science on Scaling Law Discovery domain_mixFollow99.1R2 ScoreSIMPLETES98.78898.86998.9599.031Apr 21, 2026Evaluation ResultsMethodMethodLinksR2 ScoreSIMPLETESModel=gpt-oss-120bModel=gpt-oss-120b2026.0499.1SLDAgentModel=GPT-5Model=GPT-52026.0498.8