Elamur Research⌕
☰
home / papers / 2026 / “Excuse me, may I say something...” CoLabScience, A Proactive AI Assistant for Biomedical Discovery and LLM-Expert Collaborations

“Excuse me, may I say something...” CoLabScience, A Proactive AI Assistant for Biomedical Discovery and LLM-Expert Collaborations

Paper · 2026 · Yang Wu, Jinhong Yu, Jingwei Xiong et al.

Summary

Most LLM assistants for science are reactive: they speak only when asked. In a live research meeting, that means useful remarks may never be made. This paper presents CoLabScience, an assistant that follows biomedical team discussions as they unfold and decides on its own when to contribute and what to say.

Its core is PULI (Positive-Unlabeled Learning-to-Intervene), which has three parts:

  • A reinforcement-learning coordinator sorts unlabeled dialogue turns into likely interventions and likely silences.
  • A small Observer LLM, trained with GRPO, learns the timing of interventions.
  • A larger Presenter LLM, trained with supervised fine-tuning, writes the actual suggestions.

The model's input combines the project proposal with short-term and long-term conversational memory.

For training and testing, the authors built BSDD, a set of LLM-simulated multi-role research dialogues grounded in PubMed articles. Only one intervention point per dialogue is labeled. In both simulation and human evaluation, PULI beats baselines on the timing and the usefulness of interventions, across several backbones. This suggests LLMs can act as proactive collaborators rather than passive tools.

Key points

  • BSDD (Biomedical Streaming Dialogue Dataset): simulated discussions among roles such as Pharmacologist and Clinical Physician, built from PubMed. It uses sparse positive-unlabeled labeling, with a single best intervention point per dialogue, to limit hallucinated labels.
  • Memory design: short-term memory holds the current utterance plus the two before it. Long-term memory is a running LLM summary of earlier rounds.
  • Efficiency: the lightweight Observer acts as a gate, so the expensive Presenter is called only when an intervention is needed. The rewards from both models train the coordinator end to end.
  • Evaluation:
    • Backbones were open models such as Qwen and LLaMA, compared on the same backbone and also against GPT-based baselines.
    • Six biomedical experts rated 100 sampled dialogues for dataset quality.
    • A prototype with ASR/TTS adds roughly 0.8 s of latency per turn.
  • Availability and limitations: code is at github.com/YANGWU001/CoLabScience. The authors note that evaluation has not yet covered real meetings.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓University of California, DavisDavis · United States34 works · 42,940 citations🎓Worcester Polytechnic InstituteWorcester · United States12 works · 380 citations

Also on the byline: Jiangsu University.

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksMedicine · 275 worksHealth Informatics · 214 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 3,894 works

Cite

Cite as Yang Wu, Jinhong Yu, Jingwei Xiong, Zhimin Tao and Xiaozhong Liu. 2026. “Excuse me, may I say something...” CoLabScience, A Proactive AI Assistant for Biomedical Discovery and LLM-Expert Collaborations. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.1671

@inproceedings{wu2026excuse,
  title = {{“Excuse me, may I say something...” CoLabScience, A Proactive AI Assistant for Biomedical Discovery and LLM-Expert Collaborations}},
  author = {Yang Wu, Jinhong Yu, Jingwei Xiong, Zhimin Tao and Xiaozhong Liu},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1671},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1671}
}

References

61 references, 15 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Joya Chen, Zhaoyang Lv, Shiwei Wu, Kevin Qinghong Lin, Chenan Song, Difei Gao, Jia-Wei Liu, Ziteng Gao, Dongxing Mao, and Mike Zheng Shou. 2024. Videollm-online: Online video large language model for streaming video. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio… find
  2. Juraj Gottweis, Wei-Hung Weng, Alexander Daryin, Tao Tu, Anil Palepu, Petar Sirkovic, Artiom Myaskovsky, Felix Weissenberger, Keran Rong, Ryutaro Tanno, et al. 2025. Towards an ai co-scientist. arXiv preprint arXiv:2502.18864.
  3. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, et al. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783.
  4. Amine Ben Hassouna, Hana Chaari, and Ines Belhaj. 2025. Llm-agent-umf: Llm-based agent unified modeling framework for seamless design of multi active/-passive core-agent architectures. Information Fusion, page 103865. find
  5. Kostas Hatalis, Despina Christou, Joshua Myers, Steven Jones, Keith Lambert, Adam Amos-Binks, Zohreh Dannenhauer, and Dustin Dannenhauer. 2023. Memory matters: The need to improve long-term memory in llm-agents. In Proceedings of the AAAI Symposium Series, volume 2, pages 277–280. find
  6. Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, et al. 2025. A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM Transactions on Information Systems, 43(2):…
  7. Ruofan Jin, Zaixi Zhang, Mengdi Wang, and Le Cong. 2025. Stella: Self-evolving llm agent for biomedical research. arXiv preprint arXiv:2507.02004. find
  8. John Jumper, Richard Evans, Alexander Pritzel, Tim Green, Michael Figurnov, Olaf Ronneberger, Kathryn Tunyasuvunakool, Russ Bates, Augustin Žídek, Anna Potapenko, et al. 2021. Highly accurate protein structure prediction with alphafold. nature, 596(7873):583–589.
  9. Ryuichi Kiryo, Gang Niu, Marthinus C Du Plessis, and Masashi Sugiyama. 2017. Positive-unlabeled learning with non-negative risk estimator. Advances in neural information processing systems, 30. find
  10. Ryuto Koike, Masahiro Kaneko, and Naoaki Okazaki. 2024. Outfox: Llm-generated essay detection through in-context learning with adversarially generated examples. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 21258–21266. find
  11. Shuyue Stella Li, Vidhisha Balachandran, Shangbin Feng, Jonathan Ilgen, Emma Pierson, Pang Wei Koh, and Yulia Tsvetkov. 2024. Mediq: question-asking llms for adaptive and reliable clinical reasoning. Advances in Neural Information Processing Systems. find
  12. Lizi Liao, Grace Hui Yang, and Chirag Shah. 2023. Proactive conversational agents in the post-chatgpt world. In Proceedings of the 46th international ACM SIGIR conference on research and development in information retrieval, pages 3452–3455. find
  13. Jung Hoon Lim, Sunjae Kwon, Zonghai Yao, John P Lalor, and Hong Yu. 2024. Large language model-based role-playing for personalized medical jargon extraction. arXiv preprint arXiv:2408.05555. — [arXiv:2408.05555](https://arxiv.org/abs/2408.05555) find
  14. Tianjian Liu, Hongzheng Zhao, Yuheng Liu, Xingbo Wang, and Zhenhui Peng. 2024. Compeer: A generative conversational agent for proactive peer support. In Proceedings of the 37th Annual ACM Symposium on User Interface Software and Technology, pages 1–22. find
  15. Chris Lu, Cong Lu, Robert Tjarko Lange, Jakob Foerster, Jeff Clune, and David Ha. 2024a. The ai scientist: Towards fully automated open-ended scientific discovery. arXiv preprint arXiv:2408.06292.
  16. Yaxi Lu, Shenzhi Yang, Cheng Qian, Guirong Chen, Qinyu Luo, Yesai Wu, Huadong Wang, Xin Cong, Zhong Zhang, Yankai Lin, et al. 2024b. Proactive agent: Shifting llm agents from reactive responses to active assistance. arXiv preprint arXiv:2410.12361. — [arXiv:2410.12361](https://arxiv.org/abs/2410.12361) find
  17. Zhiyong Lu, Yifan Peng, Trevor Cohen, Marzyeh Ghassemi, Chunhua Weng, and Shubo Tian. 2024c. Large language models in biomedicine and health: current research landscape and future directions. Journal of the American Medical Informatics Association, 31(9):1801–1811. find
  18. Potsawee Manakul, Adian Liusie, and Mark Gales. 2023. Selfcheckgpt: Zero-resource black-box hallucination detection for generative large language models. In Proceedings of the 2023 conference on empirical methods in natural language processing, pages 9004– 9017. find
  19. Sourab Mangrulkar, Sylvain Gugger, Lysandre Debut, Younes Belkada, Sayak Paul, Benjamin Bossan, and Marian Tietz. 2022. PEFT: State-of-the-art parameter-efficient fine-tuning methods. https://github.com/huggingface/peft. ↗
  20. Chunyu Miao, Henry Peng Zou, Yangning Li, Yankai Chen, Yibo Wang, Fangxin Wang, Yifan Li, Wooseong Yang, Bowei He, Xinni Zhang, et al. 2025. Recode-h: A benchmark for research code development with interactive human feedback. arXiv preprint arXiv:2510.06186. — [arXiv:2510.06186](https://arxiv.org/abs/2510.06186) find
  21. OpenAI. 2025. Gpt-4.1 api documentation. https://platform.openai.com/docs/models/gpt-4.1. Accessed: 2025-05-11. ↗
  22. Jing-Cheng Pang, Heng-Bo Fan, Pengyuan Wang, Jia-Hao Xiao, Nan Tang, Si-Hang Yang, Chengxing Jia, Ming-Kun Xie, Xiang Chen, Sheng-Jun Huang, et al. Interactive large language models for reliable answering under incomplete context. Transactions on Machine Learning Research. find
  23. Adam Paszke, Sam Gross, Soumith Chintala, Gregory Chanan, Edward Yang, Zachary DeVito, Zeming Lin, Alban Desmaison, Luca Antiga, and Adam Lerer. 2017. Automatic differentiation in pytorch. find
  24. Biqing Qi, Kaiyan Zhang, Kai Tian, Haoxiang Li, Zhang-Ren Chen, Sihang Zeng, Ermo Hua, Hu Jinfang, and Bowen Zhou. 2024. Large language models as biomedical hypothesis generators: A comprehensive evaluation. arXiv preprint arXiv:2407.08940.
  25. Chuan Luo, Pu Zhao, Chen Chen, Bo Qiao, Chao Du, Hongyu Zhang, Wei Wu, Shaowei Cai, Bing He, Saravanakumar Rajmohan, et al. 2021. Pulns: Positive-unlabeled learning with effective negative sample selector. In Proceedings of the AAAI conference on artificial intelligence, volume 35, pages 8784–8792. find
  26. Cheng Qian, Bingxiang He, Zhong Zhuang, Jia Deng, Yujia Qin, Xin Cong, Zhong Zhang, Jie Zhou, Yankai Lin, Zhiyuan Liu, et al. 2024. Tell me more! towards implicit user intention understanding of language model driven agents. In Proceedings of the 62nd Annual Meeting of the Association for Computatio… find
  27. Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon, and Tie-Yan Liu. 2022. Biogpt: generative pre-trained transformer for biomedical text generation and mining. Briefings in bioinformatics, 23(6):bbac409. find
  28. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. 2023. Direct preference optimization: Your language model is secretly a reward model. Advances in neural information processing systems, 36:53728–53741.
  29. Tengfei Ma, Xuan Lin, Tianle Li, Chaoyi Li, Long Chen, Peng Zhou, Xibao Cai, Xinyu Yang, Daojian Zeng, Dongsheng Cao, et al. 2024. Y-mol: a multiscale biomedical knowledge-guided large language model for drug development. arXiv preprint arXiv:2410.11550. — [arXiv:2410.11550](https://arxiv.org/abs/2410.11550) find
  30. Vishal Vivek Saley, Goonjan Saha, Rocktim Jyoti Das, Dinesh Raghu, et al. 2024. Meditod: An english dialogue dataset for medical history taking with comprehensive annotations. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 16843–16877.
  31. Eric W Sayers, Jeff Beck, Evan E Bolton, J Rodney Brister, Jessica Chan, Donald C Comeau, Ryan Connor, Michael DiCuccio, Catherine M Farrell, Michael Feldgarden, et al. 2024. Database resources of the national center for biotechnology information. Nucleic acids research, 52(D1):D33–D43. find
  32. Felix Wong, Erica J Zheng, Jacqueline A Valeri, Nina M Donghia, Melis N Anahtar, Satotaka Omori, Alicia Li, Andres Cubillos-Ruiz, Aarti Krishnan, Wengong Jin, et al. 2024. Discovery of a structural class of antibiotics with explainable deep learning. Nature, 626(7997):177–185. find
  33. Samuel Schmidgall, Yusheng Su, Ze Wang, Ximeng Sun, Jialian Wu, Xiaodong Yu, Jiang Liu, Michael Moor, Zicheng Liu, and Emad Barsoum. 2025. Agent laboratory: Using llm agents as research assistants. Findings of the Association for Computational Linguistics: EMNLP 2025, pages 5977–6043.
  34. Cheng-Kuang Wu, Zhi Rui Tam, Chao-Chung Wu, Chieh-Yen Lin, Hung-yi Lee, and Yun-Nung Chen. 2024a. I need help! evaluating llm’s ability to ask for users’ support: A case study on text-to-sql generation. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages… find
  35. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, YK Li, Yang Wu, et al. 2024. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. arXiv preprint arXiv:2402.03300.
  36. Gaurang Sriramanan, Siddhant Bharti, Vinu Sankar Sadasivan, Shoumik Saha, Priyatham Kattakinda, and Soheil Feizi. 2024. Llm-check: Investigating detection of hallucinations in large language models. Advances in Neural Information Processing Systems, 37:34188–34216. find
  37. Richard S Sutton, David McAllester, Satinder Singh, and Yishay Mansour. 1999. Policy gradient methods for reinforcement learning with function approximation. Advances in neural information processing systems, 12. find
  38. K Swanson, W Wu, NL Bulaong, JE Pak, and J Zou. 2024. The virtual lab: Ai agents design new sars-cov-2 nanobodies with experimental validation. biorxiv, 2024. URL https://www.biorxiv. org/content/early/2024/11/12/2024.11, 11. ↗
  39. Yufei Tao, Ameeta Agrawal, Judit Dombi, Tetyana Sydorenko, and Jung In Lee. 2024. Chatgpt role-play dataset: Analysis of user motives and model naturalness. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), p… find
  40. Leandro von Werra, Younes Belkada, Lewis Tunstall, Edward Beeching, Tristan Thrush, Nathan Lambert, Shengyi Huang, Kashif Rasul, and Quentin Gallouédec. 2020. TRL: Transformers Reinforcement Learning. find
  41. Hongru Wang, Rui Wang, Fei Mi, Zezhong Wang, Ruifeng Xu, and Kam-Fai Wong. 2023. Chain-of-thought prompting for responding to in-depth dialogue questions with llm. arXiv preprint arXiv:2305.11792. — [arXiv:2305.11792](https://arxiv.org/abs/2305.11792) find
  42. Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pier-ric Cistac, Tim Rault, Rémi Louf, Morgan Funtowicz, et al. 2020. Transformers: State-of-the-art natural language processing. In Proceedings of the 2020 conference on empirical methods in natural language p…
  43. Junde Wu, Jiayuan Zhu, and Yuyuan Liu. 2025a. Agentic reasoning: Reasoning llms with tools for the deep research. arXiv preprint arXiv:2502.04644. — [arXiv:2502.04644](https://arxiv.org/abs/2502.04644) find
  44. Yang Wu, Xurui Li, Xuhong Zhang, Yangyang Kang, Changlong Sun, and Xiaozhong Liu. 2023. Community-based hierarchical positive-unlabeled (pu) model fusion for chronic disease prediction. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, pages 2747–2756. find
  45. Yang Wu, Raha Moraffah, Rujing Yao, Jinhong Yu, Zhimin Tao, and Xiaozhong Liu. 2025b. Active domain knowledge acquisition with 100-dollar budget: Enhancing llms via cost-efficient, expert-involved interaction in sensitive domains. arXiv preprint arXiv:2508.17202. find
  46. Yang Wu, Chenghao Wang, Ece Gumusel, and Xiaozhong Liu. 2024b. Knowledge-infused legal wisdom: Navigating llm consultation through the lens of diagnostics and positive-unlabeled reinforcement learning. arXiv preprint arXiv:2406.03600. find
  47. Yang Wu, Rujing Yao, Tong Zhang, Yufei Shi, Zhuoren Jiang, Zhushan Li, and Xiaozhong Liu. 2025c. Teaching according to students’ aptitude: Personalized mathematics tutoring via persona-, memory-, and forgetting-aware llms. arXiv preprint arXiv:2511.15163. — [arXiv:2511.15163](https://arxiv.org/abs/2511.15163) find
  48. Yang Wu, Huayi Zhang, Yizheng Jiao, Lin Ma, Xiaozhong Liu, Jinhong Yu, Dongyu Zhang, Dezhi Yu, and Wei Xu. 2024c. Rose: A reward-oriented data selection framework for llm task-specific instruction tuning. arXiv preprint arXiv:2412.00631. find
  49. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. 2025a. Qwen3 technical report. arXiv preprint arXiv:2505.09388.
  50. Bufang Yang, Yunqi Guo, Lilin Xu, Zhenyu Yan, Hongkai Chen, Guoliang Xing, and Xiaofan Jiang. 2025b. Socialmind: Llm-based proactive ar social assistive system with human-like perception for in-situ live interactions. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologie… find
  51. Rujing Yao, Yang Wu, Chenghao Wang, Jingwei Xiong, Fang Wang, and Xiaozhong Liu. 2025a. Elevating legal llm responses: harnessing trainable logical structures and semantic knowledge with legal reasoning. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association… find
  52. Rujing Yao, Yiquan Wu, Tong Zhang, Xuhui Zhang, Yuting Huang, Yang Wu, Jiayin Yang, Changlong Sun, Fang Wang, and Xiaozhong Liu. 2025b. Intelligent legal assistant: An interactive clarification system for legal question answering. In Companion Proceedings of the ACM on Web Conference 2025, pages 293… find
  53. Denny Zhou, Nathanael Schärli, Le Hou, Jason Wei, Nathan Scales, Xuezhi Wang, Dale Schuurmans, Claire Cui, Olivier Bousquet, Quoc Le, et al. 2022. Least-to-most prompting enables complex reasoning in large language models. arXiv preprint arXiv:2205.10625.
  54. Yuyang Zhou, Guang Cheng, Kang Du, Zihan Chen, and Yuyu Zhao. 2026. Toward intelligent and secure cloud: Large language model empowered proactive defense. IEEE Communications Magazine. find
  55. Vinicius Zambaldi, David La, Alexander E Chu, Harsh-nira Patani, Amy E Danson, Tristan OC Kwan, Thomas Frerix, Rosalia G Schneider, David Saxton, Ashok Thillaisundaram, et al. 2024. De novo design of high-affinity protein binders with alphaproteo. arXiv preprint arXiv:2409.08022. — [arXiv:2409.08022](https://arxiv.org/abs/2409.08022) find
  56. Guangtao Zeng, Wenmian Yang, Zeqian Ju, Yue Yang, Sicheng Wang, Ruisi Zhang, Meng Zhou, Jiaqi Zeng, Xiangyu Dong, Ruoyu Zhang, et al. 2020. Meddialog: Large-scale medical dialogue datasets. In Proceedings of the 2020 conference on empirical methods in natural language processing (EMNLP), pages 9241–…
  57. Ceyao Zhang, Kaijie Yang, Siyi Hu, Zihao Wang, Guanghe Li, Yihang Sun, Cheng Zhang, Zhaowei Zhang, Anji Liu, Song-Chun Zhu, et al. 2024a. Proagent: building proactive cooperative agents with large language models. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 175… find
  58. Kai Zhang, Yangyang Kang, Fubang Zhao, and Xiaozhong Liu. 2024b. Llm-based medical assistant personalization with short-and long-term memory coordination. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies… find
  59. Xuan Zhang, Yang Deng, Zifeng Ren, See Kiong Ng, and Tat-Seng Chua. 2024c. Ask-before-plan: Proactive language agents for real-world planning. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 10836–10863. find
  60. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, et al. 2023. A survey of large language models. arXiv preprint arXiv:2303.18223, 1(2):1–124.
  61. Wanjun Zhong, Lianghong Guo, Qiqi Gao, He Ye, and Yanlin Wang. 2024. Memorybank: Enhancing large language models with long-term memory. In Proceedings of the AAAI conference on artificial intelligence, volume 38, pages 19724–19731. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wu-2026-excuse-me-may-i-say/