arXiv：cs.LG· Zhenchao Tang, Fang Wang, Haohuai He, Jiale Zhou, Tianxu Lv, Jun Zhu, Shouzhi Chen, Minghao Yang, Yu Wang, Jiayang Wu, Yidong Song, Yaokun Li, Jiehui Huang, Jun Zhou, Bing He, Jianhua Yao·· 7 小时前AI 评分41
用平衡微调让 LLM 对齐生物医学知识
Aligning LLMs with Biomedical Knowledge using Balanced Fine-Tuning
AI 导读
研究者提出 Balanced Fine-Tuning（BFT）双尺度后训练方法，针对生物医学文本中密集低置信度片段所编码的认知不确定性，结合组归一化 token 重加权与序列级重分配。
正文
Authors:Zhenchao Tang, Fang Wang, Haohuai He, Jiale Zhou, Tianxu Lv, Jun Zhu, Shouzhi Chen, Minghao Yang, Yu Wang, Jiayang Wu, Yidong Song, Yaokun Li, Jiehui Huang, Jun Zhou, Bing He, Jianhua Yao
Abstract:Engineering LLMs to accelerate life sciences research requires a robust alignment with biomedical knowledge. We observe that biomedical text exhibits a fundamentally different uncertainty structure from general text: dense low-confidence runs encode epistemic knowledge gaps (dense causal chains, rare entities) rather than the sparse aleatoric stylistic variation typical of general text. Based on this discovery, we propose Balanced Fine-Tuning (BFT), a dual-scale post-training method that combines group-normalized token reweighting with sequence-level reallocation toward knowledge-dense samples exhibiting dense epistemic uncertainty. Across medical evaluation, biological reasoning, sparse-reward RL, and biological representation tasks, BFT provides more consistent gains than SFT and DFT under a shared training setup. When replacing the default closed-source backbones in GeneAgent (GPT-4o) and VCWorld (Gemini-2.5-Flash), the BFT-aligned 70B model delivers stronger performance across biological process reasoning and chemical perturbation prediction. Critically, all BFT variants further improve after subsequent GRPO with sparse rewards, while SFT and DFT degrade, suggesting that epistemic-aware post-training provides a more robust policy initialization. Beyond text generation, BFT-aligned LLMs produce more accurate and professional biomedical profile texts; after encoding these profiles with a text embedding model, the resulting representations support gene-level, cell-level, and perturbation-response tasks, suggesting that BFT-enhanced generation can facilitate biological representation and, in turn, broader biomedical downstream tasks.
| Comments: | Accepted at the 40th Conference on Neural Information Processing Systems (NeurIPS 2026). Related work updated |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2511.21075 [cs.LG] |
| (or arXiv:2511.21075v4 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2511.21075 arXiv-issued DOI via DataCite |
Submission history
From: Zhenchao Tang [view email]
[v1]
Wed, 26 Nov 2025 05:34:26 UTC (5,636 KB)
[v2]
Fri, 27 Mar 2026 03:36:42 UTC (4,721 KB)
[v3]
Sun, 3 May 2026 12:36:12 UTC (6,067 KB)
[v4]
Tue, 6 Oct 2026 14:00:27 UTC (5,927 KB)
来源：arXiv：cs.LG · arxiv.org