{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:GSCN5GDX4WMLVYBMNRVVRLYPQT","short_pith_number":"pith:GSCN5GDX","canonical_record":{"source":{"id":"2002.02836","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-02-07T15:18:15Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"0f0dcd5f6dec0a0f97c3d5eb3df5218b73ed6613becd8de654a7e77a9afe41ce","abstract_canon_sha256":"20faecc573666ebc79e154eb5393172ec37a5b06dac15b2278c7d0881d576f54"},"schema_version":"1.0"},"canonical_sha256":"3484de9877e598bae02c6c6b58af0f84dcd3f2c7cb0c153c68fa6de3c7895653","source":{"kind":"arxiv","id":"2002.02836","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2002.02836","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"arxiv_version","alias_value":"2002.02836v1","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2002.02836","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"pith_short_12","alias_value":"GSCN5GDX4WML","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"pith_short_16","alias_value":"GSCN5GDX4WMLVYBM","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"pith_short_8","alias_value":"GSCN5GDX","created_at":"2026-07-05T00:38:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:GSCN5GDX4WMLVYBMNRVVRLYPQT","target":"record","payload":{"canonical_record":{"source":{"id":"2002.02836","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-02-07T15:18:15Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"0f0dcd5f6dec0a0f97c3d5eb3df5218b73ed6613becd8de654a7e77a9afe41ce","abstract_canon_sha256":"20faecc573666ebc79e154eb5393172ec37a5b06dac15b2278c7d0881d576f54"},"schema_version":"1.0"},"canonical_sha256":"3484de9877e598bae02c6c6b58af0f84dcd3f2c7cb0c153c68fa6de3c7895653","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:38:59.807387Z","signature_b64":"2IET5T4lLUUthcZUWWiHQN+0vIBF9lPX+Lg6EI6ro3me/BvP+ENnSznCAflwQZVkcmMloibXWAyBwLvNALjxAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3484de9877e598bae02c6c6b58af0f84dcd3f2c7cb0c153c68fa6de3c7895653","last_reissued_at":"2026-07-05T00:38:59.807002Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:38:59.807002Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2002.02836","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T00:38:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YrxIuAKTwhGd1en591Xhr2ReUqs6eqY1IDinbgfBSmtEMPRvzKY4ThoaD6pGTQrhfnUr6NfjgVpHsTk9DuarDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-05T15:06:23.940561Z"},"content_sha256":"4c9bdaf5b6421a89e0a4f9fe3da497cad705f4df5ac7b8885ec7ab8477c4b9b5","schema_version":"1.0","event_id":"sha256:4c9bdaf5b6421a89e0a4f9fe3da497cad705f4df5ac7b8885ec7ab8477c4b9b5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:GSCN5GDX4WMLVYBMNRVVRLYPQT","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Causally Correct Partial Models for Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Danilo J. Rezende, Fabio Viola, Frederic Besse, George Papamakarios, Hamza Merzic, Ioannis Antonoglou, Ivo Danihelka, Jane Wang, Jovana Mitrovic, Karol Gregor, Lars Buesing, Nan Rosemary Ke, Ray Jiang, Theophane Weber","submitted_at":"2020-02-07T15:18:15Z","abstract_excerpt":"In reinforcement learning, we can learn a model of future observations and rewards, and use it to plan the agent's next actions. However, jointly modeling future observations can be computationally expensive or even intractable if the observations are high-dimensional (e.g. images). For this reason, previous works have considered partial models, which model only part of the observation. In this paper, we show that partial models can be causally incorrect: they are confounded by the observations they don't model, and can therefore lead to incorrect planning. To address this, we introduce a gene"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2002.02836","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/2002.02836/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T00:38:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NJ5Rw77SgUYBwmxg2voLWlbW6hXPlYydvoZ+hx2RMh8bT+lgHBo37Z1KSzlyZApgax5YdjhiyVLormbQbJ4DAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-05T15:06:23.941657Z"},"content_sha256":"297cca64c0cf6a91773b3c0280e292bd86082802003140e694467d54f62ce2e8","schema_version":"1.0","event_id":"sha256:297cca64c0cf6a91773b3c0280e292bd86082802003140e694467d54f62ce2e8"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GSCN5GDX4WMLVYBMNRVVRLYPQT/bundle.json","state_url":"https://pith.science/pith/GSCN5GDX4WMLVYBMNRVVRLYPQT/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GSCN5GDX4WMLVYBMNRVVRLYPQT/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-05T15:06:23Z","links":{"resolver":"https://pith.science/pith/GSCN5GDX4WMLVYBMNRVVRLYPQT","bundle":"https://pith.science/pith/GSCN5GDX4WMLVYBMNRVVRLYPQT/bundle.json","state":"https://pith.science/pith/GSCN5GDX4WMLVYBMNRVVRLYPQT/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GSCN5GDX4WMLVYBMNRVVRLYPQT/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:GSCN5GDX4WMLVYBMNRVVRLYPQT","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"20faecc573666ebc79e154eb5393172ec37a5b06dac15b2278c7d0881d576f54","cross_cats_sorted":["cs.AI","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-02-07T15:18:15Z","title_canon_sha256":"0f0dcd5f6dec0a0f97c3d5eb3df5218b73ed6613becd8de654a7e77a9afe41ce"},"schema_version":"1.0","source":{"id":"2002.02836","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2002.02836","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"arxiv_version","alias_value":"2002.02836v1","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2002.02836","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"pith_short_12","alias_value":"GSCN5GDX4WML","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"pith_short_16","alias_value":"GSCN5GDX4WMLVYBM","created_at":"2026-07-05T00:38:59Z"},{"alias_kind":"pith_short_8","alias_value":"GSCN5GDX","created_at":"2026-07-05T00:38:59Z"}],"graph_snapshots":[{"event_id":"sha256:297cca64c0cf6a91773b3c0280e292bd86082802003140e694467d54f62ce2e8","target":"graph","created_at":"2026-07-05T00:38:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2002.02836/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In reinforcement learning, we can learn a model of future observations and rewards, and use it to plan the agent's next actions. However, jointly modeling future observations can be computationally expensive or even intractable if the observations are high-dimensional (e.g. images). For this reason, previous works have considered partial models, which model only part of the observation. In this paper, we show that partial models can be causally incorrect: they are confounded by the observations they don't model, and can therefore lead to incorrect planning. To address this, we introduce a gene","authors_text":"Danilo J. Rezende, Fabio Viola, Frederic Besse, George Papamakarios, Hamza Merzic, Ioannis Antonoglou, Ivo Danihelka, Jane Wang, Jovana Mitrovic, Karol Gregor, Lars Buesing, Nan Rosemary Ke, Ray Jiang, Theophane Weber","cross_cats":["cs.AI","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-02-07T15:18:15Z","title":"Causally Correct Partial Models for Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2002.02836","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:4c9bdaf5b6421a89e0a4f9fe3da497cad705f4df5ac7b8885ec7ab8477c4b9b5","target":"record","created_at":"2026-07-05T00:38:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"20faecc573666ebc79e154eb5393172ec37a5b06dac15b2278c7d0881d576f54","cross_cats_sorted":["cs.AI","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-02-07T15:18:15Z","title_canon_sha256":"0f0dcd5f6dec0a0f97c3d5eb3df5218b73ed6613becd8de654a7e77a9afe41ce"},"schema_version":"1.0","source":{"id":"2002.02836","kind":"arxiv","version":1}},"canonical_sha256":"3484de9877e598bae02c6c6b58af0f84dcd3f2c7cb0c153c68fa6de3c7895653","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3484de9877e598bae02c6c6b58af0f84dcd3f2c7cb0c153c68fa6de3c7895653","first_computed_at":"2026-07-05T00:38:59.807002Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T00:38:59.807002Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"2IET5T4lLUUthcZUWWiHQN+0vIBF9lPX+Lg6EI6ro3me/BvP+ENnSznCAflwQZVkcmMloibXWAyBwLvNALjxAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T00:38:59.807387Z","signed_message":"canonical_sha256_bytes"},"source_id":"2002.02836","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4c9bdaf5b6421a89e0a4f9fe3da497cad705f4df5ac7b8885ec7ab8477c4b9b5","sha256:297cca64c0cf6a91773b3c0280e292bd86082802003140e694467d54f62ce2e8"],"state_sha256":"835db03c44f0cfbfc1a170343f7a5ccfc7e463b7e1fb4d30633aa3d49a5550d0"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0/pLaoT5TZt6G9LwTVetAjiJFV4EIvY8uRQysNgfU0ngGyvqW0+qQMdzsaT/c58+YoR5bHAa7GfN5R8nWRndCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-05T15:06:23.944817Z","bundle_sha256":"abae85063df170f7a1a10ebfc898ca09e2dcf54d86dd625d1950b6cfc8d53beb"}}