{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:7C75GU5VWFCAWPT2DA3TPBU4LO","short_pith_number":"pith:7C75GU5V","canonical_record":{"source":{"id":"2606.04075","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-02T16:29:48Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CR","cs.CY"],"title_canon_sha256":"6b76570af319bb6981a4c57c51db6adc0bd1494bce8eb472b9f810d1e0ed84df","abstract_canon_sha256":"b6cbf023986997288764fcbbd1cf99c9c8e8a990f827af7fafc2f5d2e175dd28"},"schema_version":"1.0"},"canonical_sha256":"f8bfd353b5b1440b3e7a183737869c5b862216df2f315df79f0a5bb463ebc09a","source":{"kind":"arxiv","id":"2606.04075","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.04075","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"arxiv_version","alias_value":"2606.04075v1","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.04075","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"pith_short_12","alias_value":"7C75GU5VWFCA","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"pith_short_16","alias_value":"7C75GU5VWFCAWPT2","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"pith_short_8","alias_value":"7C75GU5V","created_at":"2026-06-04T00:06:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:7C75GU5VWFCAWPT2DA3TPBU4LO","target":"record","payload":{"canonical_record":{"source":{"id":"2606.04075","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-02T16:29:48Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CR","cs.CY"],"title_canon_sha256":"6b76570af319bb6981a4c57c51db6adc0bd1494bce8eb472b9f810d1e0ed84df","abstract_canon_sha256":"b6cbf023986997288764fcbbd1cf99c9c8e8a990f827af7fafc2f5d2e175dd28"},"schema_version":"1.0"},"canonical_sha256":"f8bfd353b5b1440b3e7a183737869c5b862216df2f315df79f0a5bb463ebc09a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-04T00:06:45.857102Z","signature_b64":"hAqn85rsXSKIUVyCmycavLgwFqJqdOZkyheBbQ9D6NxI1Gu7SKTvw14srY+Al+6H98otVZNmdiVUNBStA6l+AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f8bfd353b5b1440b3e7a183737869c5b862216df2f315df79f0a5bb463ebc09a","last_reissued_at":"2026-06-04T00:06:45.856704Z","signature_status":"signed_v1","first_computed_at":"2026-06-04T00:06:45.856704Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.04075","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":"2026-06-04T00:06:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QJnl0P7qCWOh53kjiIUF3ci/YqFL3XuPxniBm23ohfSRDVZIYP3llgwZ9DhcngCpOIaI8k+f4QRmEYJMLP1UDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-01T14:40:15.098636Z"},"content_sha256":"c39b8cfa1086b61a7699957439daee3936ea311f5ee0dd1a66db6dce8c7b03bd","schema_version":"1.0","event_id":"sha256:c39b8cfa1086b61a7699957439daee3936ea311f5ee0dd1a66db6dce8c7b03bd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:7C75GU5VWFCAWPT2DA3TPBU4LO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Large Language Models Hack Rewards, and Society","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CR","cs.CY"],"primary_cat":"cs.LG","authors_text":"Hanqi Yan, Wei Liu, Xinyi Mou, Yulan He, Zhongyu Wei","submitted_at":"2026-06-02T16:29:48Z","abstract_excerpt":"Reinforcement learning (RL) has become a dominant post-training paradigm, enabling large language models (LLMs) to learn from rewards. We observe that societal regulations are structurally similar to reward functions. They define measurable outcomes, thresholds, and exceptions, while often leaving institutional intent only partially specified. We hypothesise that the RL training process may exploit these gaps and therefore ask whether models' well-known tendency to hack reward functions during RL can scale into a more consequential failure mode named societal hacking: discovering loopholes in "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.04075","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/2606.04075/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":"2026-06-04T00:06:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SmPfXuqFS3S0+h0QZOj9J3riTQeuFEE9PXjhe6ZmrdZOrsHReN56KMgYFux6nWhLog8O6u9bH8tDExyqZeciAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-01T14:40:15.099007Z"},"content_sha256":"b69f325f61ca60d96a851baa0e83fd3e8835ffd4fd0f170e2e60aa76dcd0ebbe","schema_version":"1.0","event_id":"sha256:b69f325f61ca60d96a851baa0e83fd3e8835ffd4fd0f170e2e60aa76dcd0ebbe"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7C75GU5VWFCAWPT2DA3TPBU4LO/bundle.json","state_url":"https://pith.science/pith/7C75GU5VWFCAWPT2DA3TPBU4LO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7C75GU5VWFCAWPT2DA3TPBU4LO/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":"2026-07-01T14:40:15Z","links":{"resolver":"https://pith.science/pith/7C75GU5VWFCAWPT2DA3TPBU4LO","bundle":"https://pith.science/pith/7C75GU5VWFCAWPT2DA3TPBU4LO/bundle.json","state":"https://pith.science/pith/7C75GU5VWFCAWPT2DA3TPBU4LO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7C75GU5VWFCAWPT2DA3TPBU4LO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:7C75GU5VWFCAWPT2DA3TPBU4LO","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":"b6cbf023986997288764fcbbd1cf99c9c8e8a990f827af7fafc2f5d2e175dd28","cross_cats_sorted":["cs.AI","cs.CL","cs.CR","cs.CY"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-02T16:29:48Z","title_canon_sha256":"6b76570af319bb6981a4c57c51db6adc0bd1494bce8eb472b9f810d1e0ed84df"},"schema_version":"1.0","source":{"id":"2606.04075","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.04075","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"arxiv_version","alias_value":"2606.04075v1","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.04075","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"pith_short_12","alias_value":"7C75GU5VWFCA","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"pith_short_16","alias_value":"7C75GU5VWFCAWPT2","created_at":"2026-06-04T00:06:45Z"},{"alias_kind":"pith_short_8","alias_value":"7C75GU5V","created_at":"2026-06-04T00:06:45Z"}],"graph_snapshots":[{"event_id":"sha256:b69f325f61ca60d96a851baa0e83fd3e8835ffd4fd0f170e2e60aa76dcd0ebbe","target":"graph","created_at":"2026-06-04T00:06:45Z","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/2606.04075/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning (RL) has become a dominant post-training paradigm, enabling large language models (LLMs) to learn from rewards. We observe that societal regulations are structurally similar to reward functions. They define measurable outcomes, thresholds, and exceptions, while often leaving institutional intent only partially specified. We hypothesise that the RL training process may exploit these gaps and therefore ask whether models' well-known tendency to hack reward functions during RL can scale into a more consequential failure mode named societal hacking: discovering loopholes in ","authors_text":"Hanqi Yan, Wei Liu, Xinyi Mou, Yulan He, Zhongyu Wei","cross_cats":["cs.AI","cs.CL","cs.CR","cs.CY"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-02T16:29:48Z","title":"Large Language Models Hack Rewards, and Society"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.04075","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:c39b8cfa1086b61a7699957439daee3936ea311f5ee0dd1a66db6dce8c7b03bd","target":"record","created_at":"2026-06-04T00:06:45Z","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":"b6cbf023986997288764fcbbd1cf99c9c8e8a990f827af7fafc2f5d2e175dd28","cross_cats_sorted":["cs.AI","cs.CL","cs.CR","cs.CY"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-02T16:29:48Z","title_canon_sha256":"6b76570af319bb6981a4c57c51db6adc0bd1494bce8eb472b9f810d1e0ed84df"},"schema_version":"1.0","source":{"id":"2606.04075","kind":"arxiv","version":1}},"canonical_sha256":"f8bfd353b5b1440b3e7a183737869c5b862216df2f315df79f0a5bb463ebc09a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f8bfd353b5b1440b3e7a183737869c5b862216df2f315df79f0a5bb463ebc09a","first_computed_at":"2026-06-04T00:06:45.856704Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-04T00:06:45.856704Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hAqn85rsXSKIUVyCmycavLgwFqJqdOZkyheBbQ9D6NxI1Gu7SKTvw14srY+Al+6H98otVZNmdiVUNBStA6l+AA==","signature_status":"signed_v1","signed_at":"2026-06-04T00:06:45.857102Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.04075","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c39b8cfa1086b61a7699957439daee3936ea311f5ee0dd1a66db6dce8c7b03bd","sha256:b69f325f61ca60d96a851baa0e83fd3e8835ffd4fd0f170e2e60aa76dcd0ebbe"],"state_sha256":"3952023445affabc62c0cde052fcd87d7add45a19851fb16a5a92d356043858c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2SH0r9+r2y9cMl4/hgtnulgq20Pd4rj5nCTLpyjKdkZS1aso7mNGxEafClANuznJx4hySTHlMx4ia66d4P/bDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-01T14:40:15.100974Z","bundle_sha256":"25ba25857b7820589b440390eca4f9c0abc06f112c7933be6b50dbbfab52874f"}}