M. J. Buehler
1 indexed paper
Recent (6 mo)
1With code
0Influential cites
0Benchmarked
0Research Timeline
2026
ProjectionBench: Evaluating Scientific Hypothesis Generation in LLMs Under Progressive Information Disclosure
The paper introduces ProjectionBench, a novel benchmark that progressively discloses information to evaluate LLMs' ability to generate scientific hypotheses, demonstrating that advanced models like GPT-5.4 maintain high accuracy even with minimal context.
Highlighted terms show continued research focus across papers