{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UK63K3XRPKZWGRVJGALTEYCX7W","short_pith_number":"pith:UK63K3XR","schema_version":"1.0","canonical_sha256":"a2bdb56ef17ab36346a93017326057fdaf96fe8571b9940e6962fea5940c3a93","source":{"kind":"arxiv","id":"2606.10616","version":1},"attestation_state":"computed","paper":{"title":"Learning What to Remember: Observability-Safe Memory Retention via Constrained Optimization for Long-Horizon Language Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Kaichao Liang, Liu Mingyang, Mingxuan Yuan, Qingcan Kang, Shixiong Kai, Tao Zhong","submitted_at":"2026-06-09T09:15:33Z","abstract_excerpt":"Long-horizon language agents accumulate observations, reasoning traces, and retrieved facts that exceed their finite context windows, making memory retention a fundamental resource-allocation problem. Existing memory systems improve management through heuristic scoring, retrieval optimization, or learned compression, but largely treat retention as a local decision problem and do not explicitly model its long-term consequences under realistic observability constraints. To fill this gap, we formulate memory retention as a constrained stochastic optimization problem with explicit budget feasibili"},"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":"2606.10616","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-09T09:15:33Z","cross_cats_sorted":[],"title_canon_sha256":"ffa8b356cb4bae392613a2e402b44773445bac7824e062595a5ee473b853305e","abstract_canon_sha256":"90a20d94e0a97e0d4a8780b41df7431f51b91c37b79c531bc84fede6ba77aead"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:10:29.688878Z","signature_b64":"ShU5QILC/94/WInjazUjpG6QC9QIgyFakrM9Ys9r5vIgmB+PrZ+u6NiuUEaJm1r7IoMBnVmIHDbJPp/lmu6hBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a2bdb56ef17ab36346a93017326057fdaf96fe8571b9940e6962fea5940c3a93","last_reissued_at":"2026-06-10T01:10:29.688066Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:10:29.688066Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning What to Remember: Observability-Safe Memory Retention via Constrained Optimization for Long-Horizon Language Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Kaichao Liang, Liu Mingyang, Mingxuan Yuan, Qingcan Kang, Shixiong Kai, Tao Zhong","submitted_at":"2026-06-09T09:15:33Z","abstract_excerpt":"Long-horizon language agents accumulate observations, reasoning traces, and retrieved facts that exceed their finite context windows, making memory retention a fundamental resource-allocation problem. Existing memory systems improve management through heuristic scoring, retrieval optimization, or learned compression, but largely treat retention as a local decision problem and do not explicitly model its long-term consequences under realistic observability constraints. To fill this gap, we formulate memory retention as a constrained stochastic optimization problem with explicit budget feasibili"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10616","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.10616/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":"2606.10616","created_at":"2026-06-10T01:10:29.688168+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.10616v1","created_at":"2026-06-10T01:10:29.688168+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10616","created_at":"2026-06-10T01:10:29.688168+00:00"},{"alias_kind":"pith_short_12","alias_value":"UK63K3XRPKZW","created_at":"2026-06-10T01:10:29.688168+00:00"},{"alias_kind":"pith_short_16","alias_value":"UK63K3XRPKZWGRVJ","created_at":"2026-06-10T01:10:29.688168+00:00"},{"alias_kind":"pith_short_8","alias_value":"UK63K3XR","created_at":"2026-06-10T01:10:29.688168+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/UK63K3XRPKZWGRVJGALTEYCX7W","json":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W.json","graph_json":"https://pith.science/api/pith-number/UK63K3XRPKZWGRVJGALTEYCX7W/graph.json","events_json":"https://pith.science/api/pith-number/UK63K3XRPKZWGRVJGALTEYCX7W/events.json","paper":"https://pith.science/paper/UK63K3XR"},"agent_actions":{"view_html":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W","download_json":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W.json","view_paper":"https://pith.science/paper/UK63K3XR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.10616&json=true","fetch_graph":"https://pith.science/api/pith-number/UK63K3XRPKZWGRVJGALTEYCX7W/graph.json","fetch_events":"https://pith.science/api/pith-number/UK63K3XRPKZWGRVJGALTEYCX7W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W/action/storage_attestation","attest_author":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W/action/author_attestation","sign_citation":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W/action/citation_signature","submit_replication":"https://pith.science/pith/UK63K3XRPKZWGRVJGALTEYCX7W/action/replication_record"}},"created_at":"2026-06-10T01:10:29.688168+00:00","updated_at":"2026-06-10T01:10:29.688168+00:00"}