{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:B7FTF4IJB4DDA6HJMRDCKAL6YV","short_pith_number":"pith:B7FTF4IJ","schema_version":"1.0","canonical_sha256":"0fcb32f1090f063078e9644625017ec56533540b2fe6c758fcc92bd3e01a8d10","source":{"kind":"arxiv","id":"2106.01404","version":1},"attestation_state":"computed","paper":{"title":"Variational Empowerment as Representation Learning for Goal-Based Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Archit Sharma, Honglak Lee, Jongwook Choi, Sergey Levine, Shixiang Shane Gu","submitted_at":"2021-06-02T18:12:26Z","abstract_excerpt":"Learning to reach goal states and learning diverse skills through mutual information (MI) maximization have been proposed as principled frameworks for self-supervised reinforcement learning, allowing agents to acquire broadly applicable multitask policies with minimal reward engineering. Starting from a simple observation that the standard goal-conditioned RL (GCRL) is encapsulated by the optimization objective of variational empowerment, we discuss how GCRL and MI-based RL can be generalized into a single family of methods, which we name variational GCRL (VGCRL), interpreting variational MI m"},"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":"2106.01404","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-06-02T18:12:26Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"80e286dbea28d7a351dd520594f1ebd01eee6feaed758d7cca6ffdabce5aed49","abstract_canon_sha256":"35a2e57c1ac4726b5e7ad96a05387ff7f0d145306e28b41cec63e853aa9f634a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:45:51.004616Z","signature_b64":"MZemJP4H+YD2eFF7k/5TMSb0Fk6Uw+hIBSMIZW+NERT1gs3obdxyGB/GRkXnK0vT1CCd9lUIWtjDSEKkeO+1DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0fcb32f1090f063078e9644625017ec56533540b2fe6c758fcc92bd3e01a8d10","last_reissued_at":"2026-07-05T02:45:51.004019Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:45:51.004019Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Variational Empowerment as Representation Learning for Goal-Based Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Archit Sharma, Honglak Lee, Jongwook Choi, Sergey Levine, Shixiang Shane Gu","submitted_at":"2021-06-02T18:12:26Z","abstract_excerpt":"Learning to reach goal states and learning diverse skills through mutual information (MI) maximization have been proposed as principled frameworks for self-supervised reinforcement learning, allowing agents to acquire broadly applicable multitask policies with minimal reward engineering. Starting from a simple observation that the standard goal-conditioned RL (GCRL) is encapsulated by the optimization objective of variational empowerment, we discuss how GCRL and MI-based RL can be generalized into a single family of methods, which we name variational GCRL (VGCRL), interpreting variational MI m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.01404","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/2106.01404/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":"2106.01404","created_at":"2026-07-05T02:45:51.004090+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.01404v1","created_at":"2026-07-05T02:45:51.004090+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.01404","created_at":"2026-07-05T02:45:51.004090+00:00"},{"alias_kind":"pith_short_12","alias_value":"B7FTF4IJB4DD","created_at":"2026-07-05T02:45:51.004090+00:00"},{"alias_kind":"pith_short_16","alias_value":"B7FTF4IJB4DDA6HJ","created_at":"2026-07-05T02:45:51.004090+00:00"},{"alias_kind":"pith_short_8","alias_value":"B7FTF4IJ","created_at":"2026-07-05T02:45:51.004090+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.12261","citing_title":"Delay-Empowered Causal Hierarchical Reinforcement Learning","ref_index":31,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV","json":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV.json","graph_json":"https://pith.science/api/pith-number/B7FTF4IJB4DDA6HJMRDCKAL6YV/graph.json","events_json":"https://pith.science/api/pith-number/B7FTF4IJB4DDA6HJMRDCKAL6YV/events.json","paper":"https://pith.science/paper/B7FTF4IJ"},"agent_actions":{"view_html":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV","download_json":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV.json","view_paper":"https://pith.science/paper/B7FTF4IJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.01404&json=true","fetch_graph":"https://pith.science/api/pith-number/B7FTF4IJB4DDA6HJMRDCKAL6YV/graph.json","fetch_events":"https://pith.science/api/pith-number/B7FTF4IJB4DDA6HJMRDCKAL6YV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV/action/storage_attestation","attest_author":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV/action/author_attestation","sign_citation":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV/action/citation_signature","submit_replication":"https://pith.science/pith/B7FTF4IJB4DDA6HJMRDCKAL6YV/action/replication_record"}},"created_at":"2026-07-05T02:45:51.004090+00:00","updated_at":"2026-07-05T02:45:51.004090+00:00"}