---
title: "CollabLLM: From Passive Responders to Active Collaborators (Extended Abstract)"
authors:
  - "Shirley Wu"
  - "Michel Galley"
  - "Baolin Peng"
  - "Hao Cheng"
  - "Gavin Li"
  - "Yao Dou"
  - "Weixin Cai"
  - "James Zou"
  - "Jure Leskovec"
  - "Jianfeng Gao"
doi: "10.24963/ijcai.2026/928"
url: "https://ijcai.org/proceedings/2026/928"
pdf: "https://ijcai.org/proceedings/2026/0928.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "8289-8294"
track: "Sister Conferences Best Papers"
keywords:
  - "Machine Learning: Deep learning architectures"
  - "Natural Language Processing: Dialogue and interactive systems"
  - "Humans and AI: Human-AI collaboration"
---

# CollabLLM: From Passive Responders to Active Collaborators (Extended Abstract)

Shirley Wu, Michel Galley, Baolin Peng, Hao Cheng, Gavin Li, Yao Dou, Weixin Cai, James Zou, Jure Leskovec, Jianfeng Gao

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 8289-8294. DOI: [10.24963/ijcai.2026/928](https://doi.org/10.24963/ijcai.2026/928)

## Abstract

Large Language Models are typically trained with next-turn rewards, limiting their ability to optimize for long-term interaction. As a result, they often respond passively to ambiguous or open-ended user requests, failing to help users reach their ultimate intents and leading to inefficient conversations. To address these limitations, we introduce CollabLLM, a novel and general training framework that enhances multiturn human-LLM collaboration. Its key innovation is a collaborative simulation that estimates the long-term contribution of responses using Multiturn-aware Rewards. By reinforcement fine-tuning these rewards, CollabLLM goes beyond responding to user requests and actively uncovers user intent and offers insightful suggestions--a key step toward more human-centered AI. We also devise a multiturn interaction benchmark with three challenging tasks such as document creation. COLLABLLM significantly outperforms our baselines with averages of 18.5% higher task performance and 44.3% improved interactivity by LLM judges. Finally, we conduct a large user study with 201 judges, where CollabLLM increases user satisfaction by 17.6% and reduces user spent time by 10.4%.