Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.11462 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911588824383488 |
|---|---|
| author | Li, Xiaozhe Lyu, Tianyi Yang, Yizhao Shan, Liang Yang, Siyi Zhang, Ligao Huang, Zhuoyi Liu, Qingwen Li, Yang |
| author_facet | Li, Xiaozhe Lyu, Tianyi Yang, Yizhao Shan, Liang Yang, Siyi Zhang, Ligao Huang, Zhuoyi Liu, Qingwen Li, Yang |
| contents | Large Language Models (LLMs) struggle with long-horizon tasks due to the "context bottleneck" and the "lost-in-the-middle" phenomenon, where accumulated noise from verbose environments degrades reasoning over multi-turn interactions. To address this issue, we introduce a symbiotic framework that decouples context management from task execution. Our architecture pairs a lightweight, specialized policy model, ContextCurator, with a powerful frozen foundation model, TaskExecutor. Trained via reinforcement learning, ContextCurator actively reduces information entropy in the working memory. It aggressively prunes environmental noise while preserving reasoning anchors, that is, sparse data points that are critical for future deductions. On WebArena, our framework improves the success rate of Gemini-3.0-flash from 36.4% to 41.2% while reducing token consumption by 8.8% (from 47.4K to 43.3K). On DeepSearch, it achieves a 57.1% success rate, compared with 53.9%, while reducing token consumption by a factor of 8. Remarkably, a 7B ContextCurator matches the context management performance of GPT-4o, providing a scalable and computationally efficient paradigm for autonomous long-horizon agents. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_11462 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Escaping the Context Bottleneck: Active Context Curation for LLM Agents via Reinforcement Learning Li, Xiaozhe Lyu, Tianyi Yang, Yizhao Shan, Liang Yang, Siyi Zhang, Ligao Huang, Zhuoyi Liu, Qingwen Li, Yang Artificial Intelligence Large Language Models (LLMs) struggle with long-horizon tasks due to the "context bottleneck" and the "lost-in-the-middle" phenomenon, where accumulated noise from verbose environments degrades reasoning over multi-turn interactions. To address this issue, we introduce a symbiotic framework that decouples context management from task execution. Our architecture pairs a lightweight, specialized policy model, ContextCurator, with a powerful frozen foundation model, TaskExecutor. Trained via reinforcement learning, ContextCurator actively reduces information entropy in the working memory. It aggressively prunes environmental noise while preserving reasoning anchors, that is, sparse data points that are critical for future deductions. On WebArena, our framework improves the success rate of Gemini-3.0-flash from 36.4% to 41.2% while reducing token consumption by 8.8% (from 47.4K to 43.3K). On DeepSearch, it achieves a 57.1% success rate, compared with 53.9%, while reducing token consumption by a factor of 8. Remarkably, a 7B ContextCurator matches the context management performance of GPT-4o, providing a scalable and computationally efficient paradigm for autonomous long-horizon agents. |
| title | Escaping the Context Bottleneck: Active Context Curation for LLM Agents via Reinforcement Learning |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2604.11462 |