{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XFO6EDLVGTCMUKBUDJ723EBMKS","short_pith_number":"pith:XFO6EDLV","schema_version":"1.0","canonical_sha256":"b95de20d7534c4ca28341a7fad902c548b266635efb6079bc6caa1bbf18f1959","source":{"kind":"arxiv","id":"2403.16369","version":3},"attestation_state":"computed","paper":{"title":"Learning Action-based Representations Using Invariance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Amy Zhang, Caleb Chuck, Kevin Black, Max Rudolph, Misha Lvovsky, Scott Niekum","submitted_at":"2024-03-25T02:17:54Z","abstract_excerpt":"Robust reinforcement learning agents using high-dimensional observations must be able to identify relevant state features amidst many exogeneous distractors. A representation that captures controllability identifies these state elements by determining what affects agent control. While methods such as inverse dynamics and mutual information capture controllability for a limited number of timesteps, capturing long-horizon elements remains a challenging problem. Myopic controllability can capture the moment right before an agent crashes into a wall, but not the control-relevance of the wall while"},"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":"2403.16369","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-03-25T02:17:54Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"38abc1a67fcdf1104cc19e43995896bdde3bb0898c63317a8e2982375eafc584","abstract_canon_sha256":"682e82153bf921b8ad68e9e0cf90d6ef257f588c70453545ee6e9cd811a8bb51"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:35:44.691801Z","signature_b64":"TJiBq1ZDkhmoyE58PbpFrk4K0SE0Hm44bB8nPt7wDWCQF6kiPdqaA/MjpsfjKkECA96DUAcLw+ZHVcJJiFvOAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b95de20d7534c4ca28341a7fad902c548b266635efb6079bc6caa1bbf18f1959","last_reissued_at":"2026-07-05T08:35:44.691300Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:35:44.691300Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Action-based Representations Using Invariance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Amy Zhang, Caleb Chuck, Kevin Black, Max Rudolph, Misha Lvovsky, Scott Niekum","submitted_at":"2024-03-25T02:17:54Z","abstract_excerpt":"Robust reinforcement learning agents using high-dimensional observations must be able to identify relevant state features amidst many exogeneous distractors. A representation that captures controllability identifies these state elements by determining what affects agent control. While methods such as inverse dynamics and mutual information capture controllability for a limited number of timesteps, capturing long-horizon elements remains a challenging problem. Myopic controllability can capture the moment right before an agent crashes into a wall, but not the control-relevance of the wall while"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.16369","kind":"arxiv","version":3},"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/2403.16369/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":"2403.16369","created_at":"2026-07-05T08:35:44.691367+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.16369v3","created_at":"2026-07-05T08:35:44.691367+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.16369","created_at":"2026-07-05T08:35:44.691367+00:00"},{"alias_kind":"pith_short_12","alias_value":"XFO6EDLVGTCM","created_at":"2026-07-05T08:35:44.691367+00:00"},{"alias_kind":"pith_short_16","alias_value":"XFO6EDLVGTCMUKBU","created_at":"2026-07-05T08:35:44.691367+00:00"},{"alias_kind":"pith_short_8","alias_value":"XFO6EDLV","created_at":"2026-07-05T08:35:44.691367+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.11525","citing_title":"Learning Object Manipulation from Scratch via Contrastive Interaction","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30068","citing_title":"Predictive Objectives Discard Exogenous Control-Relevant Features: A Controlled Mechanistic Study","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS","json":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS.json","graph_json":"https://pith.science/api/pith-number/XFO6EDLVGTCMUKBUDJ723EBMKS/graph.json","events_json":"https://pith.science/api/pith-number/XFO6EDLVGTCMUKBUDJ723EBMKS/events.json","paper":"https://pith.science/paper/XFO6EDLV"},"agent_actions":{"view_html":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS","download_json":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS.json","view_paper":"https://pith.science/paper/XFO6EDLV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.16369&json=true","fetch_graph":"https://pith.science/api/pith-number/XFO6EDLVGTCMUKBUDJ723EBMKS/graph.json","fetch_events":"https://pith.science/api/pith-number/XFO6EDLVGTCMUKBUDJ723EBMKS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS/action/storage_attestation","attest_author":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS/action/author_attestation","sign_citation":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS/action/citation_signature","submit_replication":"https://pith.science/pith/XFO6EDLVGTCMUKBUDJ723EBMKS/action/replication_record"}},"created_at":"2026-07-05T08:35:44.691367+00:00","updated_at":"2026-07-05T08:35:44.691367+00:00"}