{"paper":{"title":"EXG: Self-Evolving Agents with Experience Graphs","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"EXG turns agent successes and failures into a connected graph for instant reuse across tasks.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Hanchen Wang, Lu Qin, Siyuan Zhang, Wenjie Zhang, Ying Zhang, Yuxin Jin","submitted_at":"2026-05-18T00:50:23Z","abstract_excerpt":"Large language model (LLM)-based agents have demonstrated strong capabilities in complex reasoning and problem solving through multi-step interactions, yet most deployed agents remain behaviorally static, with knowledge acquired during execution rarely translating into systematic improvement over time. In response, a growing line of work on self-evolving agents explores how agents can improve through experience during deployment, but most existing approaches either rely on ad hoc reflection limited to single-task correction or adopt unstructured memory that accumulates fragmented experience wi"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"EXG is the first experience graph designed for self-evolving agents, supporting both online, real-time graph growth during execution for immediate cross-task experience reuse, and offline reuse of a consolidated experience graph as an external memory module.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That successes and failures accumulated during agent execution can be effectively captured and related in a graph structure that enables immediate and transferable reuse without fragmentation or high overhead, as positioned against ad hoc reflection and unstructured memory in the abstract.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"EXG is an experience graph framework for self-evolving LLM agents that supports online real-time growth and offline reuse to enhance solution quality and efficiency on code generation and reasoning benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"EXG turns agent successes and failures into a connected graph for instant reuse across tasks.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a95eda3b9c47fbce5b14300f7eb3f8d82c6a29cd0697419527365f42028c7a02"},"source":{"id":"2605.17721","kind":"arxiv","version":1},"verdict":{"id":"6eaeec59-e9f3-460f-846a-f85d31164ead","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T22:14:02.846903Z","strongest_claim":"EXG is the first experience graph designed for self-evolving agents, supporting both online, real-time graph growth during execution for immediate cross-task experience reuse, and offline reuse of a consolidated experience graph as an external memory module.","one_line_summary":"EXG is an experience graph framework for self-evolving LLM agents that supports online real-time growth and offline reuse to enhance solution quality and efficiency on code generation and reasoning benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That successes and failures accumulated during agent execution can be effectively captured and related in a graph structure that enables immediate and transferable reuse without fragmentation or high overhead, as positioned against ad hoc reflection and unstructured memory in the abstract.","pith_extraction_headline":"EXG turns agent successes and failures into a connected graph for instant reuse across tasks."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.17721/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T22:31:19.378207Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T22:20:57.052636Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"shingle_duplication","ran_at":"2026-05-19T21:49:43.466689Z","status":"skipped","version":"0.1.0","findings_count":0},{"name":"citation_quote_validity","ran_at":"2026-05-19T21:49:43.296092Z","status":"skipped","version":"0.1.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.500082Z","status":"skipped","version":"1.0.0","findings_count":0},{"name":"cited_work_retraction","ran_at":"2026-05-19T21:21:57.418662Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T21:21:57.401990Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"d2ec0e0fd6c563033c9644a79d36847c319fe0539fda8a4156306b113c039074"},"references":{"count":56,"sample":[{"doi":"","year":2025,"title":"MemoryBench: A Benchmark for Memory and Continual Learning in LLM Systems","work_id":"eaff38e0-7f83-4800-aa64-f8687ace2302","ref_index":1,"cited_arxiv_id":"2510.17281","is_internal_anchor":true},{"doi":"","year":2025,"title":"FLEX: Continuous agent evolution via forward learning from experience","work_id":"20c70353-cdc5-4bb5-806a-59181ff4fbdf","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Remember Me, Refine Me: A Dynamic Procedural Memory Framework for Experience-Driven Agent Evolution","work_id":"e1808e0d-921b-46dd-b221-96110080aa71","ref_index":3,"cited_arxiv_id":"2512.10696","is_internal_anchor":true},{"doi":"","year":2021,"title":"Evaluating Large Language Models Trained on Code","work_id":"042493e9-b26f-4b4e-bbde-382072ca9b08","ref_index":4,"cited_arxiv_id":"2107.03374","is_internal_anchor":true},{"doi":"","year":null,"title":"Jizhan Fang, Xinle Deng, Haoming Xu, Ziyan Jiang, Yuqi Tang, Ziwen Xu, Shumin Deng, Yunzhi Yao, Mengru Wang, Shuofei Qiao, Huajun Chen, and Ningyu Zhang","work_id":"27b895a0-0661-4820-b6da-406212444530","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":56,"snapshot_sha256":"7d2273352019d635d4e3d85a4ae213b9ffc8d116a0dab4655f555f46d1545a24","internal_anchors":19},"formal_canon":{"evidence_count":2,"snapshot_sha256":"a3e372c84bb742fc9ee2486bfa5ed6ac65f4fb9db2f14d20072aba82e4f103e9"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}