{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:S4PJI2NXHOUQOCNLDM76GLOM72","short_pith_number":"pith:S4PJI2NX","schema_version":"1.0","canonical_sha256":"971e9469b73ba90709ab1b3fe32dccfe9c737f751fbb2a2cc88988b2a8ac5d4a","source":{"kind":"arxiv","id":"2410.00704","version":1},"attestation_state":"computed","paper":{"title":"Contrastive Abstraction for Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Elisabeth Rumetshofer, Markus Hofmarcher, Sepp Hochreiter, Vihang Patil","submitted_at":"2024-10-01T13:56:09Z","abstract_excerpt":"Learning agents with reinforcement learning is difficult when dealing with long trajectories that involve a large number of states. To address these learning problems effectively, the number of states can be reduced by abstract representations that cluster states. In principle, deep reinforcement learning can find abstract states, but end-to-end learning is unstable. We propose contrastive abstraction learning to find abstract states, where we assume that successive states in a trajectory belong to the same abstract state. Such abstract states may be basic locations, achieved subgoals, invento"},"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":"2410.00704","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-01T13:56:09Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6d70c3b84e974660d7ba22e6fc0fbe65cca7f370b932ce47889bd53a949a6e0d","abstract_canon_sha256":"68f710e886d90165a6c53bcb4a09dde00528eec4a75d3723c0e32ba044b5e763"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:14:09.230036Z","signature_b64":"IKpgV1L20HCNzN5/fR/q7b7UQ7b7C8GZbLPZP6Mj7G5HRJYZCsuIAPS+8Ydf7KW8B+EtS8YRAW08aWIIBlrVBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"971e9469b73ba90709ab1b3fe32dccfe9c737f751fbb2a2cc88988b2a8ac5d4a","last_reissued_at":"2026-07-05T09:14:09.229697Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:14:09.229697Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Contrastive Abstraction for Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Elisabeth Rumetshofer, Markus Hofmarcher, Sepp Hochreiter, Vihang Patil","submitted_at":"2024-10-01T13:56:09Z","abstract_excerpt":"Learning agents with reinforcement learning is difficult when dealing with long trajectories that involve a large number of states. To address these learning problems effectively, the number of states can be reduced by abstract representations that cluster states. In principle, deep reinforcement learning can find abstract states, but end-to-end learning is unstable. We propose contrastive abstraction learning to find abstract states, where we assume that successive states in a trajectory belong to the same abstract state. Such abstract states may be basic locations, achieved subgoals, invento"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.00704","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/2410.00704/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":"2410.00704","created_at":"2026-07-05T09:14:09.229756+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.00704v1","created_at":"2026-07-05T09:14:09.229756+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.00704","created_at":"2026-07-05T09:14:09.229756+00:00"},{"alias_kind":"pith_short_12","alias_value":"S4PJI2NXHOUQ","created_at":"2026-07-05T09:14:09.229756+00:00"},{"alias_kind":"pith_short_16","alias_value":"S4PJI2NXHOUQOCNL","created_at":"2026-07-05T09:14:09.229756+00:00"},{"alias_kind":"pith_short_8","alias_value":"S4PJI2NX","created_at":"2026-07-05T09:14:09.229756+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/S4PJI2NXHOUQOCNLDM76GLOM72","json":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72.json","graph_json":"https://pith.science/api/pith-number/S4PJI2NXHOUQOCNLDM76GLOM72/graph.json","events_json":"https://pith.science/api/pith-number/S4PJI2NXHOUQOCNLDM76GLOM72/events.json","paper":"https://pith.science/paper/S4PJI2NX"},"agent_actions":{"view_html":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72","download_json":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72.json","view_paper":"https://pith.science/paper/S4PJI2NX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.00704&json=true","fetch_graph":"https://pith.science/api/pith-number/S4PJI2NXHOUQOCNLDM76GLOM72/graph.json","fetch_events":"https://pith.science/api/pith-number/S4PJI2NXHOUQOCNLDM76GLOM72/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72/action/storage_attestation","attest_author":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72/action/author_attestation","sign_citation":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72/action/citation_signature","submit_replication":"https://pith.science/pith/S4PJI2NXHOUQOCNLDM76GLOM72/action/replication_record"}},"created_at":"2026-07-05T09:14:09.229756+00:00","updated_at":"2026-07-05T09:14:09.229756+00:00"}