Problem
Research questions and friction points this paper is trying to address.
machine-learning interatomic potentials
failure mode discovery
benchmark scores
MLIP Detective: Active Failure Mode Discovery Beyond Benchmark Scores for Machine-Learning Interatomic Potentials
🤖 AI Summary
该研究通过引入MLIP Detective框架，采用物理信息搜索方法主动发现机器学习原子间势能的隐藏失效模式，补充了基于基准测试的评估。
📝 Abstract
Universal machine-learning interatomic potentials (u-MLIPs) aim to generalize across diverse configurations. Benchmarks enable reproducible evaluation but may not expose failures outside their predefined scope. Here, we show that physics-informed search can complement benchmark-based evaluation by uncovering hidden failure modes. We introduce MLIP Detective, an agentic framework for active failure mode discovery. Starting from benchmark evidence, MLIP Detective generates falsifiable, physics-informed failure hypotheses, screens them with inexpensive simulations, and escalates only the most suspicious cases to human experts together with proposed verification protocols. Without issue-specific prompting, MLIP Detective identified and characterized a systematic anomaly in MACE-MPA-0: the model predicted some relaxed adsorbate-surface systems involving O- or F-containing adsorbates to be higher in energy than their corresponding separated fragments. Using cross-model comparisons, MLIP Detective further inferred a likely training-data origin for the anomaly, consistent with recent reports.
Innovation
Methods, ideas, or system contributions that make the work stand out.
physics-informed search
active failure mode discovery
benchmark complementation
cross-model comparisons
🔎 Similar Papers
No similar papers found.
