{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SNKKA5T4TNXHDKLQP3XABL4LMX","short_pith_number":"pith:SNKKA5T4","schema_version":"1.0","canonical_sha256":"9354a0767c9b6e71a9707eee00af8b65ffb106a6254042cf70658e326c1939a1","source":{"kind":"arxiv","id":"2606.08656","version":1},"attestation_state":"computed","paper":{"title":"From Player to Master: Enhancing Test-Time Learning of LLM Agents via Reinforcement Learning over Memory","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aohan Zeng, Can Huang, Jie Tang, Jinhua Du, Wenhan Ma, Wenxuan Huang, Xingyu Guo, Xuancheng Huang, Xu Sun, Yishuo Cai, Yuyang Hu","submitted_at":"2026-06-07T14:53:19Z","abstract_excerpt":"Large language model (LLM) agents are increasingly deployed in long-running settings where improving through experience at test time becomes important. A common approach is to update an explicit memory after each interaction to guide future decisions. However, most existing methods rely on hand-designed prompting rules, making it difficult to align memory updates with downstream objectives over multi-step horizons consistently. We propose MemoPilot, a plug-in memory copilot that explicitly trains the memory update process to improve a frozen LLM's performance across sequential interactions. We"},"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":"2606.08656","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-07T14:53:19Z","cross_cats_sorted":[],"title_canon_sha256":"05a323aa685828d4d60f932bb5ca9d2b7e68cb93dcb771e207174805747acf0e","abstract_canon_sha256":"b6186464397dd9eb2139da44255d721c751103104e0d61c7ab1c470fe898097b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:05:42.954274Z","signature_b64":"JpEVIfndgfnAsteFFHeBhpXRGWn1hGtazCyB/DIQPAs84dDpRau0KkZ9M9sbaOC3hKrjY/qNPzwglqbg1bXKBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9354a0767c9b6e71a9707eee00af8b65ffb106a6254042cf70658e326c1939a1","last_reissued_at":"2026-06-09T01:05:42.953860Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:05:42.953860Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Player to Master: Enhancing Test-Time Learning of LLM Agents via Reinforcement Learning over Memory","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aohan Zeng, Can Huang, Jie Tang, Jinhua Du, Wenhan Ma, Wenxuan Huang, Xingyu Guo, Xuancheng Huang, Xu Sun, Yishuo Cai, Yuyang Hu","submitted_at":"2026-06-07T14:53:19Z","abstract_excerpt":"Large language model (LLM) agents are increasingly deployed in long-running settings where improving through experience at test time becomes important. A common approach is to update an explicit memory after each interaction to guide future decisions. However, most existing methods rely on hand-designed prompting rules, making it difficult to align memory updates with downstream objectives over multi-step horizons consistently. We propose MemoPilot, a plug-in memory copilot that explicitly trains the memory update process to improve a frozen LLM's performance across sequential interactions. We"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.08656","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/2606.08656/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":"2606.08656","created_at":"2026-06-09T01:05:42.953918+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.08656v1","created_at":"2026-06-09T01:05:42.953918+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.08656","created_at":"2026-06-09T01:05:42.953918+00:00"},{"alias_kind":"pith_short_12","alias_value":"SNKKA5T4TNXH","created_at":"2026-06-09T01:05:42.953918+00:00"},{"alias_kind":"pith_short_16","alias_value":"SNKKA5T4TNXHDKLQ","created_at":"2026-06-09T01:05:42.953918+00:00"},{"alias_kind":"pith_short_8","alias_value":"SNKKA5T4","created_at":"2026-06-09T01:05:42.953918+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/SNKKA5T4TNXHDKLQP3XABL4LMX","json":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX.json","graph_json":"https://pith.science/api/pith-number/SNKKA5T4TNXHDKLQP3XABL4LMX/graph.json","events_json":"https://pith.science/api/pith-number/SNKKA5T4TNXHDKLQP3XABL4LMX/events.json","paper":"https://pith.science/paper/SNKKA5T4"},"agent_actions":{"view_html":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX","download_json":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX.json","view_paper":"https://pith.science/paper/SNKKA5T4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.08656&json=true","fetch_graph":"https://pith.science/api/pith-number/SNKKA5T4TNXHDKLQP3XABL4LMX/graph.json","fetch_events":"https://pith.science/api/pith-number/SNKKA5T4TNXHDKLQP3XABL4LMX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX/action/storage_attestation","attest_author":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX/action/author_attestation","sign_citation":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX/action/citation_signature","submit_replication":"https://pith.science/pith/SNKKA5T4TNXHDKLQP3XABL4LMX/action/replication_record"}},"created_at":"2026-06-09T01:05:42.953918+00:00","updated_at":"2026-06-09T01:05:42.953918+00:00"}