{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:CLRCJLR66G75ZEXEBT3GQQKXE4","short_pith_number":"pith:CLRCJLR6","schema_version":"1.0","canonical_sha256":"12e224ae3ef1bfdc92e40cf6684157273dbd5b1f2bfa0c2173fcbfa5b32099b4","source":{"kind":"arxiv","id":"1912.01217","version":2},"attestation_state":"computed","paper":{"title":"SafeLife 1.0: Exploring Side Effects in Complex Environments","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Carroll L. Wainwright, Peter Eckersley","submitted_at":"2019-12-03T06:44:48Z","abstract_excerpt":"We present SafeLife, a publicly available reinforcement learning environment that tests the safety of reinforcement learning agents. It contains complex, dynamic, tunable, procedurally generated levels with many opportunities for unsafe behavior. Agents are graded both on their ability to maximize their explicit reward and on their ability to operate safely without unnecessary side effects. We train agents to maximize rewards using proximal policy optimization and score them on a suite of benchmark levels. The resulting agents are performant but not safe -- they tend to cause large side effect"},"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":"1912.01217","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2019-12-03T06:44:48Z","cross_cats_sorted":[],"title_canon_sha256":"0ccd928e943847f1ada8737b9612238b9a2e2d862475c58d6c595e68429c6198","abstract_canon_sha256":"6f170a5128e215a774da0829eca285e2dd1d2a670e14495df707f644ee7fdd4a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:18:27.974441Z","signature_b64":"b77qbyGH9z7+ny0H5WOPqUqnFGq23JpkPMteNT0aGDhand9svzVdxoLXLD4fJgqKt2gIYml8Z3K6l/8GH4xCBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"12e224ae3ef1bfdc92e40cf6684157273dbd5b1f2bfa0c2173fcbfa5b32099b4","last_reissued_at":"2026-07-05T02:18:27.973986Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:18:27.973986Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SafeLife 1.0: Exploring Side Effects in Complex Environments","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Carroll L. Wainwright, Peter Eckersley","submitted_at":"2019-12-03T06:44:48Z","abstract_excerpt":"We present SafeLife, a publicly available reinforcement learning environment that tests the safety of reinforcement learning agents. It contains complex, dynamic, tunable, procedurally generated levels with many opportunities for unsafe behavior. Agents are graded both on their ability to maximize their explicit reward and on their ability to operate safely without unnecessary side effects. We train agents to maximize rewards using proximal policy optimization and score them on a suite of benchmark levels. The resulting agents are performant but not safe -- they tend to cause large side effect"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.01217","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/1912.01217/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":"1912.01217","created_at":"2026-07-05T02:18:27.974045+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.01217v2","created_at":"2026-07-05T02:18:27.974045+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.01217","created_at":"2026-07-05T02:18:27.974045+00:00"},{"alias_kind":"pith_short_12","alias_value":"CLRCJLR66G75","created_at":"2026-07-05T02:18:27.974045+00:00"},{"alias_kind":"pith_short_16","alias_value":"CLRCJLR66G75ZEXE","created_at":"2026-07-05T02:18:27.974045+00:00"},{"alias_kind":"pith_short_8","alias_value":"CLRCJLR6","created_at":"2026-07-05T02:18:27.974045+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2109.13916","citing_title":"Unsolved Problems in ML Safety","ref_index":195,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4","json":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4.json","graph_json":"https://pith.science/api/pith-number/CLRCJLR66G75ZEXEBT3GQQKXE4/graph.json","events_json":"https://pith.science/api/pith-number/CLRCJLR66G75ZEXEBT3GQQKXE4/events.json","paper":"https://pith.science/paper/CLRCJLR6"},"agent_actions":{"view_html":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4","download_json":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4.json","view_paper":"https://pith.science/paper/CLRCJLR6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.01217&json=true","fetch_graph":"https://pith.science/api/pith-number/CLRCJLR66G75ZEXEBT3GQQKXE4/graph.json","fetch_events":"https://pith.science/api/pith-number/CLRCJLR66G75ZEXEBT3GQQKXE4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4/action/storage_attestation","attest_author":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4/action/author_attestation","sign_citation":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4/action/citation_signature","submit_replication":"https://pith.science/pith/CLRCJLR66G75ZEXEBT3GQQKXE4/action/replication_record"}},"created_at":"2026-07-05T02:18:27.974045+00:00","updated_at":"2026-07-05T02:18:27.974045+00:00"}