{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:U7DTLTNJL7VPVCUHNCO33U4MRE","short_pith_number":"pith:U7DTLTNJ","canonical_record":{"source":{"id":"1805.08328","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-05-22T00:14:32Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"6fc076d655857dba9b724a7df5bc9f53f4f65260c62c7281559020028805f67c","abstract_canon_sha256":"ad1a54b3524f9e2d8e41ddd072cb92cb1dcdeea8f71063bbbf352f2f32370e4a"},"schema_version":"1.0"},"canonical_sha256":"a7c735cda95feafa8a87689dbdd38c892a6a327ce33b99d4c3869bc833e24bf3","source":{"kind":"arxiv","id":"1805.08328","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1805.08328","created_at":"2026-05-17T23:55:38Z"},{"alias_kind":"arxiv_version","alias_value":"1805.08328v2","created_at":"2026-05-17T23:55:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1805.08328","created_at":"2026-05-17T23:55:38Z"},{"alias_kind":"pith_short_12","alias_value":"U7DTLTNJL7VP","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_16","alias_value":"U7DTLTNJL7VPVCUH","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_8","alias_value":"U7DTLTNJ","created_at":"2026-05-18T12:32:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:U7DTLTNJL7VPVCUHNCO33U4MRE","target":"record","payload":{"canonical_record":{"source":{"id":"1805.08328","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-05-22T00:14:32Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"6fc076d655857dba9b724a7df5bc9f53f4f65260c62c7281559020028805f67c","abstract_canon_sha256":"ad1a54b3524f9e2d8e41ddd072cb92cb1dcdeea8f71063bbbf352f2f32370e4a"},"schema_version":"1.0"},"canonical_sha256":"a7c735cda95feafa8a87689dbdd38c892a6a327ce33b99d4c3869bc833e24bf3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:55:38.656008Z","signature_b64":"IQB+QhOjpfAmnXWoSulgotxx0ExRzRTFK6XIvf+ZAoUogT8WIBEbHLHdLQcKLnAQINnZ7fzkugiv0FtQINTuBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a7c735cda95feafa8a87689dbdd38c892a6a327ce33b99d4c3869bc833e24bf3","last_reissued_at":"2026-05-17T23:55:38.655485Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:55:38.655485Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1805.08328","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":"2026-05-17T23:55:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4kFlyRN9JX+zDz3RJBgIopnHtMpRXx3z0tLGk+3h1DmRm2Mz/TsW9CLvfzS1nOnZ41qDbiYA9QjskjWBeZLsAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T09:43:43.367841Z"},"content_sha256":"70a20226c469b2e76f5a3e523dbab91f0214991c97eeca9b816122e2cfa2285e","schema_version":"1.0","event_id":"sha256:70a20226c469b2e76f5a3e523dbab91f0214991c97eeca9b816122e2cfa2285e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:U7DTLTNJL7VPVCUHNCO33U4MRE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Verifiable Reinforcement Learning via Policy Extraction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Armando Solar-Lezama, Osbert Bastani, Yewen Pu","submitted_at":"2018-05-22T00:14:32Z","abstract_excerpt":"While deep reinforcement learning has successfully solved many challenging control tasks, its real-world applicability has been limited by the inability to ensure the safety of learned policies. We propose an approach to verifiable reinforcement learning by training decision tree policies, which can represent complex policies (since they are nonparametric), yet can be efficiently verified using existing techniques (since they are highly structured). The challenge is that decision tree policies are difficult to train. We propose VIPER, an algorithm that combines ideas from model compression and"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1805.08328","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":""},"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":"2026-05-17T23:55:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EK4SMc+1DjdNVujoY/HiOHv0Vw+nJcTZQj9PdKmlZhK0hthzGQddHpDpxSPGl/E/cvxYkrxmCKgGWrLcadDlBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T09:43:43.368175Z"},"content_sha256":"cbc0625abd00c017672a92a69924e84aef285fd59566d513b8b3fb3c72351f56","schema_version":"1.0","event_id":"sha256:cbc0625abd00c017672a92a69924e84aef285fd59566d513b8b3fb3c72351f56"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/U7DTLTNJL7VPVCUHNCO33U4MRE/bundle.json","state_url":"https://pith.science/pith/U7DTLTNJL7VPVCUHNCO33U4MRE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/U7DTLTNJL7VPVCUHNCO33U4MRE/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":"2026-05-28T09:43:43Z","links":{"resolver":"https://pith.science/pith/U7DTLTNJL7VPVCUHNCO33U4MRE","bundle":"https://pith.science/pith/U7DTLTNJL7VPVCUHNCO33U4MRE/bundle.json","state":"https://pith.science/pith/U7DTLTNJL7VPVCUHNCO33U4MRE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/U7DTLTNJL7VPVCUHNCO33U4MRE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:U7DTLTNJL7VPVCUHNCO33U4MRE","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":"ad1a54b3524f9e2d8e41ddd072cb92cb1dcdeea8f71063bbbf352f2f32370e4a","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-05-22T00:14:32Z","title_canon_sha256":"6fc076d655857dba9b724a7df5bc9f53f4f65260c62c7281559020028805f67c"},"schema_version":"1.0","source":{"id":"1805.08328","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1805.08328","created_at":"2026-05-17T23:55:38Z"},{"alias_kind":"arxiv_version","alias_value":"1805.08328v2","created_at":"2026-05-17T23:55:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1805.08328","created_at":"2026-05-17T23:55:38Z"},{"alias_kind":"pith_short_12","alias_value":"U7DTLTNJL7VP","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_16","alias_value":"U7DTLTNJL7VPVCUH","created_at":"2026-05-18T12:32:56Z"},{"alias_kind":"pith_short_8","alias_value":"U7DTLTNJ","created_at":"2026-05-18T12:32:56Z"}],"graph_snapshots":[{"event_id":"sha256:cbc0625abd00c017672a92a69924e84aef285fd59566d513b8b3fb3c72351f56","target":"graph","created_at":"2026-05-17T23:55:38Z","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"},"paper":{"abstract_excerpt":"While deep reinforcement learning has successfully solved many challenging control tasks, its real-world applicability has been limited by the inability to ensure the safety of learned policies. We propose an approach to verifiable reinforcement learning by training decision tree policies, which can represent complex policies (since they are nonparametric), yet can be efficiently verified using existing techniques (since they are highly structured). The challenge is that decision tree policies are difficult to train. We propose VIPER, an algorithm that combines ideas from model compression and","authors_text":"Armando Solar-Lezama, Osbert Bastani, Yewen Pu","cross_cats":["stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-05-22T00:14:32Z","title":"Verifiable Reinforcement Learning via Policy Extraction"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1805.08328","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":"sha256:70a20226c469b2e76f5a3e523dbab91f0214991c97eeca9b816122e2cfa2285e","target":"record","created_at":"2026-05-17T23:55:38Z","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":"ad1a54b3524f9e2d8e41ddd072cb92cb1dcdeea8f71063bbbf352f2f32370e4a","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-05-22T00:14:32Z","title_canon_sha256":"6fc076d655857dba9b724a7df5bc9f53f4f65260c62c7281559020028805f67c"},"schema_version":"1.0","source":{"id":"1805.08328","kind":"arxiv","version":2}},"canonical_sha256":"a7c735cda95feafa8a87689dbdd38c892a6a327ce33b99d4c3869bc833e24bf3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a7c735cda95feafa8a87689dbdd38c892a6a327ce33b99d4c3869bc833e24bf3","first_computed_at":"2026-05-17T23:55:38.655485Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:55:38.655485Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IQB+QhOjpfAmnXWoSulgotxx0ExRzRTFK6XIvf+ZAoUogT8WIBEbHLHdLQcKLnAQINnZ7fzkugiv0FtQINTuBA==","signature_status":"signed_v1","signed_at":"2026-05-17T23:55:38.656008Z","signed_message":"canonical_sha256_bytes"},"source_id":"1805.08328","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:70a20226c469b2e76f5a3e523dbab91f0214991c97eeca9b816122e2cfa2285e","sha256:cbc0625abd00c017672a92a69924e84aef285fd59566d513b8b3fb3c72351f56"],"state_sha256":"2bf2b4ee71a6b4f9f4ce3d71250832783c7072b26c2534818c3f37df67e80f75"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"c4eqJShViZVUoBEDPmWmT6xaebsOSF9213MYQimaZcPcwbOqMjbOOMKTjmE1oo/O324wHG7FoN03eRLtgIfxCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-28T09:43:43.370328Z","bundle_sha256":"5ac0141fad32936f4c1a2a6c0edfd355a2b8560dacd4b0f851586fbe122572de"}}