{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZLUVBY7X3RIUVFP7IIY6ERV5TE","short_pith_number":"pith:ZLUVBY7X","schema_version":"1.0","canonical_sha256":"cae950e3f7dc514a95ff4231e246bd9929f93c9f12bd9c060e47a7305cd5a0f4","source":{"kind":"arxiv","id":"2503.03660","version":4},"attestation_state":"computed","paper":{"title":"Chunking the Critic: A Transformer-based Soft Actor-Critic with N-Step Returns","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dong Tian, Gerhard Neumann, Onur Celik","submitted_at":"2025-03-05T16:47:36Z","abstract_excerpt":"We introduce a sequence-conditioned critic for Soft Actor-Critic (SAC) that models trajectory context with a lightweight Transformer and trains on aggregated $N$-step targets. Unlike prior approaches that (i) score state-action pairs in isolation or (ii) rely on actor-side action chunking to handle long horizons, our method strengthens the critic itself by conditioning on short trajectory segments and integrating multi-step returns -- without importance sampling (IS). The resulting sequence-aware value estimates capture the critical temporal structure for extended-horizon and sparse-reward pro"},"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":"2503.03660","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-03-05T16:47:36Z","cross_cats_sorted":[],"title_canon_sha256":"d85d09b85ea10b4ff3ea6932385c4d40a957da37436d6b1aaaa788e22b6130c5","abstract_canon_sha256":"bb22d7184e254517ce0d054372a9ab4c399918d71c0c2cb75a4b63804bcc9620"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-08T01:03:43.719998Z","signature_b64":"AUdINjTtq+Xs38tlRyDFgxbC9BfKhsDnBoe04xDe1sm9IkjsJaq2IEEh2St/JJpGVTqWiKeoqtAB/cxj5v90Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cae950e3f7dc514a95ff4231e246bd9929f93c9f12bd9c060e47a7305cd5a0f4","last_reissued_at":"2026-06-08T01:03:43.719115Z","signature_status":"signed_v1","first_computed_at":"2026-06-08T01:03:43.719115Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Chunking the Critic: A Transformer-based Soft Actor-Critic with N-Step Returns","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dong Tian, Gerhard Neumann, Onur Celik","submitted_at":"2025-03-05T16:47:36Z","abstract_excerpt":"We introduce a sequence-conditioned critic for Soft Actor-Critic (SAC) that models trajectory context with a lightweight Transformer and trains on aggregated $N$-step targets. Unlike prior approaches that (i) score state-action pairs in isolation or (ii) rely on actor-side action chunking to handle long horizons, our method strengthens the critic itself by conditioning on short trajectory segments and integrating multi-step returns -- without importance sampling (IS). The resulting sequence-aware value estimates capture the critical temporal structure for extended-horizon and sparse-reward pro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.03660","kind":"arxiv","version":4},"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/2503.03660/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":"2503.03660","created_at":"2026-06-08T01:03:43.719257+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.03660v4","created_at":"2026-06-08T01:03:43.719257+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.03660","created_at":"2026-06-08T01:03:43.719257+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZLUVBY7X3RIU","created_at":"2026-06-08T01:03:43.719257+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZLUVBY7X3RIUVFP7","created_at":"2026-06-08T01:03:43.719257+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZLUVBY7X","created_at":"2026-06-08T01:03:43.719257+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2507.07969","citing_title":"Reinforcement Learning with Action Chunking","ref_index":75,"is_internal_anchor":true},{"citing_arxiv_id":"2605.05544","citing_title":"Adaptive Q-Chunking for Offline-to-Online Reinforcement Learning","ref_index":51,"is_internal_anchor":true},{"citing_arxiv_id":"2604.12509","citing_title":"Whole-Body Mobile Manipulation using Offline Reinforcement Learning on Sub-optimal Controllers","ref_index":50,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE","json":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE.json","graph_json":"https://pith.science/api/pith-number/ZLUVBY7X3RIUVFP7IIY6ERV5TE/graph.json","events_json":"https://pith.science/api/pith-number/ZLUVBY7X3RIUVFP7IIY6ERV5TE/events.json","paper":"https://pith.science/paper/ZLUVBY7X"},"agent_actions":{"view_html":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE","download_json":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE.json","view_paper":"https://pith.science/paper/ZLUVBY7X","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.03660&json=true","fetch_graph":"https://pith.science/api/pith-number/ZLUVBY7X3RIUVFP7IIY6ERV5TE/graph.json","fetch_events":"https://pith.science/api/pith-number/ZLUVBY7X3RIUVFP7IIY6ERV5TE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE/action/storage_attestation","attest_author":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE/action/author_attestation","sign_citation":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE/action/citation_signature","submit_replication":"https://pith.science/pith/ZLUVBY7X3RIUVFP7IIY6ERV5TE/action/replication_record"}},"created_at":"2026-06-08T01:03:43.719257+00:00","updated_at":"2026-06-08T01:03:43.719257+00:00"}