{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:EG5H5ITLNWRIGIH2Y42SXAG22P","short_pith_number":"pith:EG5H5ITL","schema_version":"1.0","canonical_sha256":"21ba7ea26b6da28320fac7352b80dad3fc61aa86e92c70106787316da78efc3e","source":{"kind":"arxiv","id":"2605.23089","version":1},"attestation_state":"computed","paper":{"title":"Dreaming Smoothly and Sample Efficiently with Gradient Penalized Latent Dynamics","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"P. R. Kumar (1) ((1) Texas A&M University), Romil V. Sonigra (1)","submitted_at":"2026-05-21T22:40:40Z","abstract_excerpt":"Model-based reinforcement learning improves sample efficiency by learning a world model. However, existing latent world models such as DreamerV3 do not explicitly enforce local smoothness in their learned transition dynamics, leaving a useful inductive bias for transition dynamics learning unexploited. We propose GPLD, a gradient-penalized latent dynamics regularizer for DreamerV3 that applies a row-wise Jacobian penalty to the posterior latent distribution to encourage locally smooth transition learning. We show that this penalty can be interpreted as the continuous-latent analog of finite-di"},"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":"2605.23089","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-21T22:40:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"325e84393b2012ecd44208f326748fbca0d6c6ec2c6790fdf0c4250c6bc3d9d5","abstract_canon_sha256":"e001e9cf035356bf6e6cd484991785777e8fdb885b62db886cde2d4069fcf2a8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:37.847161Z","signature_b64":"1DZZ0rZN3d8xRsxZ0dA8x5mVcOEP71zKamcAJcc9BudrMJ08/PiP2Rnf/x37eBpQYoeNEfktlcNDPdHSPXfSCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"21ba7ea26b6da28320fac7352b80dad3fc61aa86e92c70106787316da78efc3e","last_reissued_at":"2026-05-25T02:01:37.846398Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:37.846398Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dreaming Smoothly and Sample Efficiently with Gradient Penalized Latent Dynamics","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"P. R. Kumar (1) ((1) Texas A&M University), Romil V. Sonigra (1)","submitted_at":"2026-05-21T22:40:40Z","abstract_excerpt":"Model-based reinforcement learning improves sample efficiency by learning a world model. However, existing latent world models such as DreamerV3 do not explicitly enforce local smoothness in their learned transition dynamics, leaving a useful inductive bias for transition dynamics learning unexploited. We propose GPLD, a gradient-penalized latent dynamics regularizer for DreamerV3 that applies a row-wise Jacobian penalty to the posterior latent distribution to encourage locally smooth transition learning. We show that this penalty can be interpreted as the continuous-latent analog of finite-di"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.23089","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/2605.23089/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":"2605.23089","created_at":"2026-05-25T02:01:37.846527+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.23089v1","created_at":"2026-05-25T02:01:37.846527+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.23089","created_at":"2026-05-25T02:01:37.846527+00:00"},{"alias_kind":"pith_short_12","alias_value":"EG5H5ITLNWRI","created_at":"2026-05-25T02:01:37.846527+00:00"},{"alias_kind":"pith_short_16","alias_value":"EG5H5ITLNWRIGIH2","created_at":"2026-05-25T02:01:37.846527+00:00"},{"alias_kind":"pith_short_8","alias_value":"EG5H5ITL","created_at":"2026-05-25T02:01:37.846527+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/EG5H5ITLNWRIGIH2Y42SXAG22P","json":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P.json","graph_json":"https://pith.science/api/pith-number/EG5H5ITLNWRIGIH2Y42SXAG22P/graph.json","events_json":"https://pith.science/api/pith-number/EG5H5ITLNWRIGIH2Y42SXAG22P/events.json","paper":"https://pith.science/paper/EG5H5ITL"},"agent_actions":{"view_html":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P","download_json":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P.json","view_paper":"https://pith.science/paper/EG5H5ITL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.23089&json=true","fetch_graph":"https://pith.science/api/pith-number/EG5H5ITLNWRIGIH2Y42SXAG22P/graph.json","fetch_events":"https://pith.science/api/pith-number/EG5H5ITLNWRIGIH2Y42SXAG22P/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P/action/storage_attestation","attest_author":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P/action/author_attestation","sign_citation":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P/action/citation_signature","submit_replication":"https://pith.science/pith/EG5H5ITLNWRIGIH2Y42SXAG22P/action/replication_record"}},"created_at":"2026-05-25T02:01:37.846527+00:00","updated_at":"2026-05-25T02:01:37.846527+00:00"}