arXiv cs.AI· Yifei Wang, Botian Wang, YiFei Wang, Jiangtao Feng, Xinbo Zhang, Yawen Ouyang, Yixin Wu, Lihao Wang, Xiaoqing Zheng, Xuanjing Huang, Hao Zhou·· 12 小时前AI 评分37
Solvent：面向多步逆合成的长程推理 8B 模型
Solvent: Long-Horizon Reasoning for Multi-Step Retrosynthesis
AI 导读
8B 模型 Solvent 将长程推理扩展到多步逆合成，通过两轮后训练课程学习交替进行迭代合成规划、外部可得性验证与路线自适应修正。经 RL 带来 2.1 倍性能提升后，Solvent 在四个基准上平均得分 61.86%，超过中等规模语言模型和此前基于搜索的逆合成系统，并可媲美前沿万亿参数 LLM。
正文 · 原文
Authors:Yifei Wang, Botian Wang, YiFei Wang, Jiangtao Feng, Xinbo Zhang, Yawen Ouyang, Yixin Wu, Lihao Wang, Xiaoqing Zheng, Xuanjing Huang, Hao Zhou
Abstract: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 $2.1\times$ 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.
| Comments: | 49 pages, 18 figures. Code: this https URL ; Model: this https URL |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.32308 [cs.AI] |
| (or arXiv:2610.32308v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.32308 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Wang Yifei [view email]
[v1]
Fri, 23 Oct 2026 08:07:43 UTC (1,887 KB)
来源：arXiv cs.AI · arxiv.org