{
  "abstract": "Transition-state searches remain a major bottleneck in reaction discovery, as identifying valid saddle-point structures requires numerous expensive quantum-chemical calculations. Generative models can reduce this burden by proposing candidates from reactant and product geometries, but supervised training on geometries alone does not enforce the physical conditions required of a transition state, while enforcing them during sampling is costly. We introduce Trèfle, a one-step flow model that amortizes physics-guided generation into training: a mean-flow objective replaces multi-step sampling, a low-cost surrogate replaces density functional theory (DFT) reward evaluations, and reinforcement-learning post-training with rewards that enforce those conditions replaces inference-time guidance. Trèfle generates accurate transition states at $30\\times$ the throughput of multi-step baselines. It transfers few-shot to ten unseen transition metals and generalizes zero-shot to molecules far larger than any in training. With DFT refinement and reaction-path validation, Trèfle-seeded searches recover all 13 unseen $γ$-ketohydroperoxide channels, while post-training raises per-attempt recovery from 28% to 50% across 85 unseen bimolecular channels. Trèfle thus provides fast, reliable seeds for automated reaction discovery.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2609.32073",
  "author_total": 6,
  "authors": [
    {
      "name": "Yunyang Li"
    },
    {
      "name": "Zechang Sun"
    },
    {
      "name": "Kuang Yu"
    },
    {
      "name": "Wen Yan"
    },
    {
      "name": "Mark Gerstein"
    },
    {
      "name": "Hung Q. Pham"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/d5gba160nhm21.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2609.32073",
  "fulltext_url": null,
  "id": "d5gba160nhm21",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2609.32073",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "physics.chem-ph"
    ],
    [
      "openalex",
      "2505|Machine Learning in Materials Science"
    ]
  ],
  "openalex": "W7214718381",
  "pdf_url": "https://arxiv.org/pdf/2609.32073",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-25",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/d5gba160nhm21.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:25:12Z",
      "native_id": "2609.32073",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2609.32073"
    }
  ],
  "subjects": [
    {
      "field": "chemistry",
      "origin": "arxiv:physics.chem-ph",
      "subfield": "physical"
    }
  ],
  "title": "Reinforcement learning amortizes transition-state physics into one-step flow models",
  "title_text": "Reinforcement learning amortizes transition-state physics into one-step flow models",
  "type": "preprint",
  "updated_at": "2026-10-30T12:25:12Z",
  "url": "https://research.draconitic.com/paper/d5gba160nhm21",
  "venue": "arXiv",
  "versions": []
}