ICA: Information-Aware Credit Assignment for Visually Grounded Long-Horizon Information-Seeking Agents

[2602.10863] ICA: Information-Aware Credit Assignment for Visually Grounded Long-Horizon Information-Seeking Agents Skip to main content Search Submit Donate Log in Computer Science > Machine Learning arXiv:2602.10863 (cs) [Submitted on 11 Feb 2026 ( v1 ), last revised 25 Aug 2026 (this version, v2)] Title: ICA: Information-Aware Credit Assignment for Visually Grounded Long-Horizon Information-Seeking Agents Authors: Cong Pang , Xuyu Feng , Yujie Yi , Jiaqi Su , Zixuan Chen , Jiawei Hong , Tiankuo Yao , Nang Yuan , Jiapeng Luo , Lewei Lu , Xin Lou View PDF HTML (experimental) Abstract: Long-horizon reinforcement learning for information seeking agents remains difficult because terminal rewards reveal whether the final answer is correct, but not which acquired information enabled it. This difficulty is amplified by text-derived webpage observations, where parsing, truncation, and summarization often produce incomplete and unstable content representations across trajectories. We propose an evidence-centric framework for web agent learning that represents information acquired through tools as identifiable units for comparison across trajectories. In particular, fetched webpages are represented as rendered snapshots, preserving layout and multimodal content as stable content-level observations. Building on these units, we introduce Information-Aware Credit Assignment, a post hoc reward propagation method that estimates turn-level utility scores from rollout success rates and assigns dense rewards to intermediate steps that introduced high-utility information. Integrated with GSPO, our method consistently improves performance on BrowseComp, GAIA, Xbench-DS, and Seal-0. Code and datasets will be released at this https URL . Comments: Accepted to the Main Conference of EMNLP 2026 Subjects: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI) Cite as: arXiv:2602.10863 [cs.LG] (or arXiv:2602.10863v2 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2602.10863 Focus to learn more arXiv-issued DOI via DataCite Journal reference: Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pp. 7541-7556, 2026 Submission history From: Cong Pang [ view email ] [v1] Wed, 11 Feb 2026 13:50:19 UTC (1,736 KB) [v2] Tue, 25 Aug 2026 06:38:30 UTC (3,182 KB) Full-text links: Access Paper: View PDF HTML (experimental) TeX Source view license Additional Features Audio Summary Current browse context: cs.LG < prev | next > new | recent | 2026-02 Change to browse by: cs cs.AI References & Citations SciX Google Scholar Semantic Scholar export BibTeX citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer ( What is the Explorer? ) Connected Papers Toggle Connected Papers ( What is Connected Papers? ) Litmaps Toggle Litmaps ( What is Litmaps? ) scite.ai Toggle scite Smart Citations ( What are Smart Citations? ) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv ( What is alphaXiv? ) Links to Code Toggle CatalyzeX Code Finder for Papers ( What is CatalyzeX? ) DagsHub Toggle DagsHub ( What is DagsHub? ) GotitPub Toggle Gotit.pub ( What is GotitPub? ) Huggingface Toggle Hugging Face ( What is Huggingface? ) ScienceCast Toggle ScienceCast ( What is ScienceCast? ) Demos Demos Replicate Toggle Replicate ( What is Replicate? ) Spaces Toggle Hugging Face Spaces ( What is Spaces? ) Spaces Toggle TXYZ.AI ( What is TXYZ.AI? ) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower ( What are Influence Flowers? ) Core recommender toggle CORE Recommender ( What is CORE? ) IArxiv recommender toggle IArxiv Recommender ( What is IArxiv? ) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs . Which authors of this paper are endorsers? | Disable MathJax ( What is MathJax? ) We gratefully acknowledge support from our major funders , member institutions , and all contributors. About · Help · Contact · Subscribe · Copyright · Privacy · Accessibility · Operational Status (opens in new tab)

Related documents

• ICA: Information-Aware Credit Assignment for Visually Grounded Long-Horizon Information-Seeking Agents#670489
• Robust image representation for classification, retrieval and object discovery#1166399
• A Comprehensive Review of Explainable Artificial Intelligence (XAI) in Computer Vision#1067519
• Why explain visual experience in terms of content?#759382
• XAI-Arena: Can LLMs Assess the Quality of XAI Explanations?#987176