{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:7FI66XKCSV4Y6OLWI4JWUJAHFI","short_pith_number":"pith:7FI66XKC","canonical_record":{"source":{"id":"2202.04849","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-02-10T05:43:41Z","cross_cats_sorted":[],"title_canon_sha256":"e161008c60bbc0fb97e51efdc21c5e15c28cea6955b3c2c0099b8ee56eea7766","abstract_canon_sha256":"eda7a1c4504eb8510fea95f63f7fb138da8116bd5fa48ab552d0e354af1de0be"},"schema_version":"1.0"},"canonical_sha256":"f951ef5d4295798f397647136a24072a2f929b3b7506e1449a9b8d473b666e39","source":{"kind":"arxiv","id":"2202.04849","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2202.04849","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"arxiv_version","alias_value":"2202.04849v2","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.04849","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"pith_short_12","alias_value":"7FI66XKCSV4Y","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"pith_short_16","alias_value":"7FI66XKCSV4Y6OLW","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"pith_short_8","alias_value":"7FI66XKC","created_at":"2026-07-05T04:36:34Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:7FI66XKCSV4Y6OLWI4JWUJAHFI","target":"record","payload":{"canonical_record":{"source":{"id":"2202.04849","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-02-10T05:43:41Z","cross_cats_sorted":[],"title_canon_sha256":"e161008c60bbc0fb97e51efdc21c5e15c28cea6955b3c2c0099b8ee56eea7766","abstract_canon_sha256":"eda7a1c4504eb8510fea95f63f7fb138da8116bd5fa48ab552d0e354af1de0be"},"schema_version":"1.0"},"canonical_sha256":"f951ef5d4295798f397647136a24072a2f929b3b7506e1449a9b8d473b666e39","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:36:34.926364Z","signature_b64":"1vwyWYcaT1KGeoEXSfWBYuifCob1LB/U2pKKsBeog4aHxZciIsrMPxcIPW9amdKwIiuMua/YzaNS78N/hHyeBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f951ef5d4295798f397647136a24072a2f929b3b7506e1449a9b8d473b666e39","last_reissued_at":"2026-07-05T04:36:34.925888Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:36:34.925888Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2202.04849","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T04:36:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HdgSu/umytA7kCt+ChZ4qNKJsLNBMy8QiHd3Gwv9OuU3vMNzlPSc66EEWfhNUS3ES7rapa1TtJ9rtP5UVa2oBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T00:25:19.252488Z"},"content_sha256":"8c1a7e62527fb63c8789c73db691ad9c9691cfa9c2b43f9c7c1970c16cf8ab76","schema_version":"1.0","event_id":"sha256:8c1a7e62527fb63c8789c73db691ad9c9691cfa9c2b43f9c7c1970c16cf8ab76"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:7FI66XKCSV4Y6OLWI4JWUJAHFI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SAFER: Data-Efficient and Safe Reinforcement Learning via Skill Acquisition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Dai, Dylan Slack, Nevan Wichers, Yinlam Chow","submitted_at":"2022-02-10T05:43:41Z","abstract_excerpt":"Methods that extract policy primitives from offline demonstrations using deep generative models have shown promise at accelerating reinforcement learning(RL) for new tasks. Intuitively, these methods should also help to trainsafeRLagents because they enforce useful skills. However, we identify these techniques are not well equipped for safe policy learning because they ignore negative experiences(e.g., unsafe or unsuccessful), focusing only on positive experiences, which harms their ability to generalize to new tasks safely. Rather, we model the latentsafetycontextusing principled contrastive "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.04849","kind":"arxiv","version":2},"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/2202.04849/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":"2026-07-05T04:36:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"U2JPHsitt0Nmq8D9TYMMIakZtDH6N3Rpssj1fAnKaRx/8OYKEocibJojNF/1kAEMgcm5sMCpTYg8U/DLH0p2AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T00:25:19.252987Z"},"content_sha256":"86aae17c3fe21d11fb39704e0451fee3a42014540e048fb3999847550ae87aaf","schema_version":"1.0","event_id":"sha256:86aae17c3fe21d11fb39704e0451fee3a42014540e048fb3999847550ae87aaf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7FI66XKCSV4Y6OLWI4JWUJAHFI/bundle.json","state_url":"https://pith.science/pith/7FI66XKCSV4Y6OLWI4JWUJAHFI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7FI66XKCSV4Y6OLWI4JWUJAHFI/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":"2026-08-07T00:25:19Z","links":{"resolver":"https://pith.science/pith/7FI66XKCSV4Y6OLWI4JWUJAHFI","bundle":"https://pith.science/pith/7FI66XKCSV4Y6OLWI4JWUJAHFI/bundle.json","state":"https://pith.science/pith/7FI66XKCSV4Y6OLWI4JWUJAHFI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7FI66XKCSV4Y6OLWI4JWUJAHFI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:7FI66XKCSV4Y6OLWI4JWUJAHFI","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":"eda7a1c4504eb8510fea95f63f7fb138da8116bd5fa48ab552d0e354af1de0be","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-02-10T05:43:41Z","title_canon_sha256":"e161008c60bbc0fb97e51efdc21c5e15c28cea6955b3c2c0099b8ee56eea7766"},"schema_version":"1.0","source":{"id":"2202.04849","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2202.04849","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"arxiv_version","alias_value":"2202.04849v2","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.04849","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"pith_short_12","alias_value":"7FI66XKCSV4Y","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"pith_short_16","alias_value":"7FI66XKCSV4Y6OLW","created_at":"2026-07-05T04:36:34Z"},{"alias_kind":"pith_short_8","alias_value":"7FI66XKC","created_at":"2026-07-05T04:36:34Z"}],"graph_snapshots":[{"event_id":"sha256:86aae17c3fe21d11fb39704e0451fee3a42014540e048fb3999847550ae87aaf","target":"graph","created_at":"2026-07-05T04:36:34Z","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/2202.04849/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Methods that extract policy primitives from offline demonstrations using deep generative models have shown promise at accelerating reinforcement learning(RL) for new tasks. Intuitively, these methods should also help to trainsafeRLagents because they enforce useful skills. However, we identify these techniques are not well equipped for safe policy learning because they ignore negative experiences(e.g., unsafe or unsuccessful), focusing only on positive experiences, which harms their ability to generalize to new tasks safely. Rather, we model the latentsafetycontextusing principled contrastive ","authors_text":"Bo Dai, Dylan Slack, Nevan Wichers, Yinlam Chow","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-02-10T05:43:41Z","title":"SAFER: Data-Efficient and Safe Reinforcement Learning via Skill Acquisition"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.04849","kind":"arxiv","version":2},"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":"sha256:8c1a7e62527fb63c8789c73db691ad9c9691cfa9c2b43f9c7c1970c16cf8ab76","target":"record","created_at":"2026-07-05T04:36:34Z","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":"eda7a1c4504eb8510fea95f63f7fb138da8116bd5fa48ab552d0e354af1de0be","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-02-10T05:43:41Z","title_canon_sha256":"e161008c60bbc0fb97e51efdc21c5e15c28cea6955b3c2c0099b8ee56eea7766"},"schema_version":"1.0","source":{"id":"2202.04849","kind":"arxiv","version":2}},"canonical_sha256":"f951ef5d4295798f397647136a24072a2f929b3b7506e1449a9b8d473b666e39","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f951ef5d4295798f397647136a24072a2f929b3b7506e1449a9b8d473b666e39","first_computed_at":"2026-07-05T04:36:34.925888Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:36:34.925888Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1vwyWYcaT1KGeoEXSfWBYuifCob1LB/U2pKKsBeog4aHxZciIsrMPxcIPW9amdKwIiuMua/YzaNS78N/hHyeBw==","signature_status":"signed_v1","signed_at":"2026-07-05T04:36:34.926364Z","signed_message":"canonical_sha256_bytes"},"source_id":"2202.04849","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8c1a7e62527fb63c8789c73db691ad9c9691cfa9c2b43f9c7c1970c16cf8ab76","sha256:86aae17c3fe21d11fb39704e0451fee3a42014540e048fb3999847550ae87aaf"],"state_sha256":"fe5612776ef8f16ef93ad012627018f0b1d4166541ee7dd9dc0ffebbce7db8a1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4PCWenF77OFgSI/wCbz2OxYDLetC748T9V4GZ+CVvkZTfZZkjfkQHiL9SgHRn2rVOG0oGympqaWoghJR6SNDCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T00:25:19.262158Z","bundle_sha256":"3d161556a95e5e193dc59e302cdb9880ec6e180ddf58bc22ded1bfef4edcc341"}}