arXiv Machine Learning By Can Polat, Mustafa Kurban, Erchin Serpedin, Hasan Kurban

Grounded verification of chemical and materials reasoning: detection is the bottleneck

Read the original on arXiv Machine Learning →

arXiv:2607. 17417v1 Announce Type: new Abstract: Large language models confabulate chemical objects (molecular formulas, space groups, formation energies) in fluent reasoning traces, concentrated on long-tail entities where confidence is least trustworthy.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

Hugging Face Trending Papers
Jul 19

Grounded verification of chemical and materials reasoning: detection is the bottleneck

Large language models confabulate chemical objects (molecular formulas, space groups, formation energies) in fluent reasoning traces, concentrated on long-tail entities where confidence is least trustworthy. Deterministic, database-grounded verification can catch and repair such errors without the coverage cost of blanket retrieval; the binding constraint, we find, is detection, not repair.