arxiv.org·2026/10/26
为什么重要 展示了 LLM 长程推理与 RL 后训练在分子逆合成任务上的能力，对 AI4Science 工程有参考价值。
多步逆合成需要将目标分子通过连续反应步骤递归转化为可购买的构建块，近期工作开始借助大语言模型（LLM）推理来应对这一任务。在其最新演进中，LLM 推理已进入长程推理阶段：模型在相互依赖的步骤间追求目标、与环境交互，并适应反馈。
这一转变显著提升了 LLM 能力，AI 驱动的数学研究突破即为例证。在这一前沿，我们提出 Solvent，一个 8B LLM，将长程推理成功扩展到多步逆合成。
通过两轮后训练课程，Solvent 学会交替进行迭代合成规划、外部可用性验证和自适应路线修订。实验表明，在 RL 带来 2.1倍 性能提升后，Solvent 在四个基准上得分 61.86%，超过中等规模语言模型和先前基于搜索的逆合成系统，并可与前沿万亿参数 LLM 相媲美。
Multi-step retrosynthesis requires recursively transforming a target molecule through successive reaction steps into purchasable building blocks, and recent work has begun to draw on large language model (LLM) reasoning to tackle the task.
In its latest evolution, LLM reasoning has entered the long-horizon regime: models pursue objectives across interdependent steps, interact with environments, and adapt to feedback.
This shift has markedly advanced LLM capabilities, as illustrated by breakthroughs in AI-driven mathematical research.
At this frontier, we introduce Solvent, an 8B LLM that successfully extends long-horizon reasoning to multi-step retrosynthesis.
Through a two-round post-training curriculum, Solvent learns to interleave iterative synthesis planning, external availability verification, and adaptive route revision.
Experiments show that, following a 2.1× performance boost from RL, Solvent scores 61.86% across four benchmarks, surpassing mid-scale language models and prior search-based retrosynthesis systems, and rivaling frontier trillion-parameter LLMs.
原版排版与图表 · 手机无法内嵌时可在浏览器打开。PDF 插件内选文不能传给网页助手。