{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:PW5WZJPZSDHTTQL6XK74IJEHPB","short_pith_number":"pith:PW5WZJPZ","schema_version":"1.0","canonical_sha256":"7dbb6ca5f990cf39c17ebabfc42487787944cd982772d00289dc6888b2d0f72d","source":{"kind":"arxiv","id":"2312.13327","version":6},"attestation_state":"computed","paper":{"title":"In-Context Reinforcement Learning for Variable Action Spaces","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexander Nikulin, Ilya Zisman, Sergey Kolesnikov, Viacheslav Sinii, Vladislav Kurenkov","submitted_at":"2023-12-20T16:58:55Z","abstract_excerpt":"Recently, it has been shown that transformers pre-trained on diverse datasets with multi-episode contexts can generalize to new reinforcement learning tasks in-context. A key limitation of previously proposed models is their reliance on a predefined action space size and structure. The introduction of a new action space often requires data re-collection and model re-training, which can be costly for some applications. In our work, we show that it is possible to mitigate this issue by proposing the Headless-AD model that, despite being trained only once, is capable of generalizing to discrete 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":"2312.13327","kind":"arxiv","version":6},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-12-20T16:58:55Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"4516161db04067e898934ded53997f7e772403d9d17a77252714bb4face4ec1e","abstract_canon_sha256":"481741e548063d8741fb7bf6a4f96556448cd1fb4c83638338ec00109261a152"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:38:35.725859Z","signature_b64":"eobpxEXWQHInEU+ORlxL7MrtwZkjtlWOupDFBTFRAPNGXOrSO6/IzMfWTFVXAeEyHWc0cObITk3rqy6nK54BAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7dbb6ca5f990cf39c17ebabfc42487787944cd982772d00289dc6888b2d0f72d","last_reissued_at":"2026-07-05T08:38:35.725369Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:38:35.725369Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"In-Context Reinforcement Learning for Variable Action Spaces","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexander Nikulin, Ilya Zisman, Sergey Kolesnikov, Viacheslav Sinii, Vladislav Kurenkov","submitted_at":"2023-12-20T16:58:55Z","abstract_excerpt":"Recently, it has been shown that transformers pre-trained on diverse datasets with multi-episode contexts can generalize to new reinforcement learning tasks in-context. A key limitation of previously proposed models is their reliance on a predefined action space size and structure. The introduction of a new action space often requires data re-collection and model re-training, which can be costly for some applications. In our work, we show that it is possible to mitigate this issue by proposing the Headless-AD model that, despite being trained only once, is capable of generalizing to discrete a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.13327","kind":"arxiv","version":6},"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/2312.13327/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":"2312.13327","created_at":"2026-07-05T08:38:35.725428+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.13327v6","created_at":"2026-07-05T08:38:35.725428+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.13327","created_at":"2026-07-05T08:38:35.725428+00:00"},{"alias_kind":"pith_short_12","alias_value":"PW5WZJPZSDHT","created_at":"2026-07-05T08:38:35.725428+00:00"},{"alias_kind":"pith_short_16","alias_value":"PW5WZJPZSDHTTQL6","created_at":"2026-07-05T08:38:35.725428+00:00"},{"alias_kind":"pith_short_8","alias_value":"PW5WZJPZ","created_at":"2026-07-05T08:38:35.725428+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.16509","citing_title":"Learning-Based Sparsification of Dynamic Graphs in Robotic Exploration Algorithms","ref_index":30,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB","json":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB.json","graph_json":"https://pith.science/api/pith-number/PW5WZJPZSDHTTQL6XK74IJEHPB/graph.json","events_json":"https://pith.science/api/pith-number/PW5WZJPZSDHTTQL6XK74IJEHPB/events.json","paper":"https://pith.science/paper/PW5WZJPZ"},"agent_actions":{"view_html":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB","download_json":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB.json","view_paper":"https://pith.science/paper/PW5WZJPZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.13327&json=true","fetch_graph":"https://pith.science/api/pith-number/PW5WZJPZSDHTTQL6XK74IJEHPB/graph.json","fetch_events":"https://pith.science/api/pith-number/PW5WZJPZSDHTTQL6XK74IJEHPB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB/action/storage_attestation","attest_author":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB/action/author_attestation","sign_citation":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB/action/citation_signature","submit_replication":"https://pith.science/pith/PW5WZJPZSDHTTQL6XK74IJEHPB/action/replication_record"}},"created_at":"2026-07-05T08:38:35.725428+00:00","updated_at":"2026-07-05T08:38:35.725428+00:00"}