{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:XLZTRT52DFIFRZP6ICYFW72Y46","short_pith_number":"pith:XLZTRT52","schema_version":"1.0","canonical_sha256":"baf338cfba195058e5fe40b05b7f58e7a88476312353e31d2913a86b38c0e34f","source":{"kind":"arxiv","id":"2604.11811","version":2},"attestation_state":"computed","paper":{"title":"M$^\\star$: Every Task Deserves Its Own Memory Harness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"LLM agents perform better with task-specific memory programs evolved as Python code than with any fixed shared design.","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.PL","authors_text":"Mirror Xu, Shiwei Zhang, Shujie Liu, Wanlu Shi, Wenbo Pan, Xiangyang Zhou, Xiaohua Jia","submitted_at":"2026-04-10T03:22:26Z","abstract_excerpt":"Large language model agents rely on specialized memory systems to accumulate and reuse knowledge during extended interactions. Recent architectures typically adopt a fixed memory design tailored to specific domains, such as semantic retrieval for conversations or skills reused for coding. However, a memory system optimized for one purpose frequently fails to transfer to others. To address this limitation, we introduce M$^\\star$, a method that automatically discovers task-optimized memory harnesses through executable program evolution. Specifically, M$^\\star$ models an agent memory system as a "},"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":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2604.11811","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.PL","submitted_at":"2026-04-10T03:22:26Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"5940f3f552839e945d0d9d51ae0c4950f5ad766c0b0911d674649645196452be","abstract_canon_sha256":"615e6f1e42d13bb7870ed92f3f47a8849dbf1aea0366307132a79d9d740043ce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:29.940959Z","signature_b64":"W9sDJI9woTzYOkXwQ5unXa3PXvTrpomuDW7A8DCa5XBjTNPoNguQwhElQqy9h0h1yZLMQQIKNPwY9LxpiD6/Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"baf338cfba195058e5fe40b05b7f58e7a88476312353e31d2913a86b38c0e34f","last_reissued_at":"2026-05-26T01:03:29.940052Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:29.940052Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"M$^\\star$: Every Task Deserves Its Own Memory Harness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"LLM agents perform better with task-specific memory programs evolved as Python code than with any fixed shared design.","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.PL","authors_text":"Mirror Xu, Shiwei Zhang, Shujie Liu, Wanlu Shi, Wenbo Pan, Xiangyang Zhou, Xiaohua Jia","submitted_at":"2026-04-10T03:22:26Z","abstract_excerpt":"Large language model agents rely on specialized memory systems to accumulate and reuse knowledge during extended interactions. Recent architectures typically adopt a fixed memory design tailored to specific domains, such as semantic retrieval for conversations or skills reused for coding. However, a memory system optimized for one purpose frequently fails to transfer to others. To address this limitation, we introduce M$^\\star$, a method that automatically discovers task-optimized memory harnesses through executable program evolution. Specifically, M$^\\star$ models an agent memory system as a "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"M$^* improves performance over existing fixed-memory baselines robustly across all evaluated tasks. Furthermore, the evolved memory programs exhibit structurally distinct processing mechanisms for each domain.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That reflective population-based code evolution on the chosen benchmarks will reliably discover memory programs that are both superior and generalizable beyond the specific evaluation tasks and failure modes analyzed.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"M* evolves distinct Python memory programs per task via population-based reflective search, outperforming fixed-memory baselines on conversation, planning, and reasoning benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"LLM agents perform better with task-specific memory programs evolved as Python code than with any fixed shared design.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"feac4ba2381661eef6410860175d7320b30380b7196a390f4c7caeb7e37d229e"},"source":{"id":"2604.11811","kind":"arxiv","version":2},"verdict":{"id":"492efbc2-7f1c-4924-aa07-ace7911b35ea","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T17:34:40.321222Z","strongest_claim":"M$^* improves performance over existing fixed-memory baselines robustly across all evaluated tasks. Furthermore, the evolved memory programs exhibit structurally distinct processing mechanisms for each domain.","one_line_summary":"M* evolves distinct Python memory programs per task via population-based reflective search, outperforming fixed-memory baselines on conversation, planning, and reasoning benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That reflective population-based code evolution on the chosen benchmarks will reliably discover memory programs that are both superior and generalizable beyond the specific evaluation tasks and failure modes analyzed.","pith_extraction_headline":"LLM agents perform better with task-specific memory programs evolved as Python code than with any fixed shared design."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.11811/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","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":"2604.11811","created_at":"2026-05-26T01:03:29.940158+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.11811v2","created_at":"2026-05-26T01:03:29.940158+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.11811","created_at":"2026-05-26T01:03:29.940158+00:00"},{"alias_kind":"pith_short_12","alias_value":"XLZTRT52DFIF","created_at":"2026-05-26T01:03:29.940158+00:00"},{"alias_kind":"pith_short_16","alias_value":"XLZTRT52DFIFRZP6","created_at":"2026-05-26T01:03:29.940158+00:00"},{"alias_kind":"pith_short_8","alias_value":"XLZTRT52","created_at":"2026-05-26T01:03:29.940158+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/XLZTRT52DFIFRZP6ICYFW72Y46","json":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46.json","graph_json":"https://pith.science/api/pith-number/XLZTRT52DFIFRZP6ICYFW72Y46/graph.json","events_json":"https://pith.science/api/pith-number/XLZTRT52DFIFRZP6ICYFW72Y46/events.json","paper":"https://pith.science/paper/XLZTRT52"},"agent_actions":{"view_html":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46","download_json":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46.json","view_paper":"https://pith.science/paper/XLZTRT52","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.11811&json=true","fetch_graph":"https://pith.science/api/pith-number/XLZTRT52DFIFRZP6ICYFW72Y46/graph.json","fetch_events":"https://pith.science/api/pith-number/XLZTRT52DFIFRZP6ICYFW72Y46/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46/action/storage_attestation","attest_author":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46/action/author_attestation","sign_citation":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46/action/citation_signature","submit_replication":"https://pith.science/pith/XLZTRT52DFIFRZP6ICYFW72Y46/action/replication_record"}},"created_at":"2026-05-26T01:03:29.940158+00:00","updated_at":"2026-05-26T01:03:29.940158+00:00"}