{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4GRNZWZPMNCEH2VEYYDJZL7OUU","short_pith_number":"pith:4GRNZWZP","schema_version":"1.0","canonical_sha256":"e1a2dcdb2f634443eaa4c6069cafeea53f8cc8a6f9e73c1fcbd16f4da22d3000","source":{"kind":"arxiv","id":"2605.12978","version":1},"attestation_state":"computed","paper":{"title":"Useful Memories Become Faulty When Continuously Updated by LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Consolidated memories from LLMs degrade over repeated updates and can perform worse than using no memory at all.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bingxuan Li, Dianqi Li, Dylan Zhang, Hao Peng, Yanshan Lin, Yihang Sun, Zhengkun Wu","submitted_at":"2026-05-13T04:15:50Z","abstract_excerpt":"Learning from past experience benefits from two complementary forms of memory: episodic traces -- raw trajectories of what happened -- and consolidated abstractions distilled across many episodes into reusable, schema-like lessons. Recent agentic-memory systems pursue the consolidated form: an LLM rewrites past trajectories into a textual memory bank that it continuously updates with new interactions, promising self-improving agents without parameter updates. Yet we find that such consolidated memories produced by today's LLMs are often faulty even when derived from useful experiences. As cons"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":true,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.12978","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T04:15:50Z","cross_cats_sorted":[],"title_canon_sha256":"8d110404a9f0bd78130fa0f993d6c99bcd8b4b87b34ffa6f80d90d055fc4044e","abstract_canon_sha256":"2f7370ef81a575c6e0e9d975fcbb16718b09c785774323377c7f21157b6ab71a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:09:08.679863Z","signature_b64":"VdPqXWLUoLmqScxqM9BN+jVioEv8I0sz17wvaq0+rhSm/l+q+5yk0AOizUDQ1TzGGdYBJkOvc/TsIR5nTXCWCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e1a2dcdb2f634443eaa4c6069cafeea53f8cc8a6f9e73c1fcbd16f4da22d3000","last_reissued_at":"2026-05-18T03:09:08.679222Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:09:08.679222Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Useful Memories Become Faulty When Continuously Updated by LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Consolidated memories from LLMs degrade over repeated updates and can perform worse than using no memory at all.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bingxuan Li, Dianqi Li, Dylan Zhang, Hao Peng, Yanshan Lin, Yihang Sun, Zhengkun Wu","submitted_at":"2026-05-13T04:15:50Z","abstract_excerpt":"Learning from past experience benefits from two complementary forms of memory: episodic traces -- raw trajectories of what happened -- and consolidated abstractions distilled across many episodes into reusable, schema-like lessons. Recent agentic-memory systems pursue the consolidated form: an LLM rewrites past trajectories into a textual memory bank that it continuously updates with new interactions, promising self-improving agents without parameter updates. Yet we find that such consolidated memories produced by today's LLMs are often faulty even when derived from useful experiences. As cons"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Consolidated memories produced by today's LLMs are often faulty even when derived from useful experiences. As consolidation proceeds, memory utility first rises, then degrades, and can fall below the no-memory baseline. More surprisingly, even when consolidating from ground-truth solutions, GPT-5.4 fails on 54% of a set of ARC-AGI problems it had previously solved without memory.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the observed degradation is caused by the consolidation step itself rather than by limitations specific to the tested models, tasks, or update schedules, and that the ARC-AGI Stream environment sufficiently represents real-world agent memory use.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"LLM-consolidated memories in agents degrade over continuous updates even from useful experiences, causing up to 54% failure on previously solved ARC-AGI problems, while episodic retention preserves accuracy.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Consolidated memories from LLMs degrade over repeated updates and can perform worse than using no memory at all.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7a2d1b4cf8e8611f38624aaeaec705568f76cdad810c25524fb70f9bd1f3036c"},"source":{"id":"2605.12978","kind":"arxiv","version":1},"verdict":{"id":"37294837-d99e-4c75-b111-c076633e814c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:55:18.938103Z","strongest_claim":"Consolidated memories produced by today's LLMs are often faulty even when derived from useful experiences. As consolidation proceeds, memory utility first rises, then degrades, and can fall below the no-memory baseline. More surprisingly, even when consolidating from ground-truth solutions, GPT-5.4 fails on 54% of a set of ARC-AGI problems it had previously solved without memory.","one_line_summary":"LLM-consolidated memories in agents degrade over continuous updates even from useful experiences, causing up to 54% failure on previously solved ARC-AGI problems, while episodic retention preserves accuracy.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the observed degradation is caused by the consolidation step itself rather than by limitations specific to the tested models, tasks, or update schedules, and that the ARC-AGI Stream environment sufficiently represents real-world agent memory use.","pith_extraction_headline":"Consolidated memories from LLMs degrade over repeated updates and can perform worse than using no memory at all."},"references":{"count":15,"sample":[{"doi":"10.1145/3586183.3606763","year":2016,"title":"URLhttps://arxiv.org/abs/2511.00162. Morris Moscovitch, Roberto Cabeza, Gordon Winocur, and Lynn Nadel. Episodic memory and beyond: The hippocampus and neocortex in transformation.Annual Review of Psy","work_id":"b5bf85fe-4fb7-4966-b0b2-9ccf9d3b11b9","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"You may RETAIN entries by index, MERGE several into a cleaner entry, or DROP entries by omitting them from the output","work_id":"74ad0fe5-09d4-41c0-8a7a-a317c557a031","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"from_existing","work_id":"3ca905f6-4a57-4724-b2be-8811c86c14c6","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"When to use: The task has two same-sized input grids and the output has the same height but double the width, arranged as a left-right concatenation. The left half reproduces the shape pattern from th","work_id":"a2e258e5-061d-4f28-8bd5-cbea6ce00bbe","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"reason\" in your reply). You MUST pick one existing strategy -- no other action is accepted: B) **Use an existing strategy**: {","work_id":"51e1acea-5313-482f-830e-6b37b2cdad5a","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":15,"snapshot_sha256":"2876651a7b979671b5e7ee5ec270103ff09a39b4bec308dbea1e6cf8d1dae610","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.12978","created_at":"2026-05-18T03:09:08.679334+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.12978v1","created_at":"2026-05-18T03:09:08.679334+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12978","created_at":"2026-05-18T03:09:08.679334+00:00"},{"alias_kind":"pith_short_12","alias_value":"4GRNZWZPMNCE","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"4GRNZWZPMNCEH2VE","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"4GRNZWZP","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU","json":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU.json","graph_json":"https://pith.science/api/pith-number/4GRNZWZPMNCEH2VEYYDJZL7OUU/graph.json","events_json":"https://pith.science/api/pith-number/4GRNZWZPMNCEH2VEYYDJZL7OUU/events.json","paper":"https://pith.science/paper/4GRNZWZP"},"agent_actions":{"view_html":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU","download_json":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU.json","view_paper":"https://pith.science/paper/4GRNZWZP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.12978&json=true","fetch_graph":"https://pith.science/api/pith-number/4GRNZWZPMNCEH2VEYYDJZL7OUU/graph.json","fetch_events":"https://pith.science/api/pith-number/4GRNZWZPMNCEH2VEYYDJZL7OUU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU/action/storage_attestation","attest_author":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU/action/author_attestation","sign_citation":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU/action/citation_signature","submit_replication":"https://pith.science/pith/4GRNZWZPMNCEH2VEYYDJZL7OUU/action/replication_record"}},"created_at":"2026-05-18T03:09:08.679334+00:00","updated_at":"2026-05-18T03:09:08.679334+00:00"}