Jul 2026
Before the Action: Benchmarking LLMs on Prospective Hypothesis Discovery
Results support treating PHD as a distinct target for evaluating how LLMs formulate investigative directions when final conclusions are withheld, with gains for several lower-performing models on HypoArena but regressions for other systems, including a top-performing model.
Tianyun Zhong, Wangyi Jiang, Wei Wang et al.
· arXiv.org · 0 citations