{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:24DYVYK2EG4N2V2F5Y55TCSKCW","short_pith_number":"pith:24DYVYK2","canonical_record":{"source":{"id":"2511.21667","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-26T18:42:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"be28bb4727885b6009197b0f861d9872d1dad43c25333ac91ceb7f0a93f27334","abstract_canon_sha256":"02bb3829985d2c84f4c6359ecc4d8f148409c51f23f53a28db65aeae38b51dbc"},"schema_version":"1.0"},"canonical_sha256":"d7078ae15a21b8dd5745ee3bd98a4a15af6ce99e4da1bbd6cd57ed131425bbd4","source":{"kind":"arxiv","id":"2511.21667","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2511.21667","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"arxiv_version","alias_value":"2511.21667v4","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.21667","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"pith_short_12","alias_value":"24DYVYK2EG4N","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"pith_short_16","alias_value":"24DYVYK2EG4N2V2F","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"pith_short_8","alias_value":"24DYVYK2","created_at":"2026-06-05T01:14:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:24DYVYK2EG4N2V2F5Y55TCSKCW","target":"record","payload":{"canonical_record":{"source":{"id":"2511.21667","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-26T18:42:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"be28bb4727885b6009197b0f861d9872d1dad43c25333ac91ceb7f0a93f27334","abstract_canon_sha256":"02bb3829985d2c84f4c6359ecc4d8f148409c51f23f53a28db65aeae38b51dbc"},"schema_version":"1.0"},"canonical_sha256":"d7078ae15a21b8dd5745ee3bd98a4a15af6ce99e4da1bbd6cd57ed131425bbd4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:14:31.302389Z","signature_b64":"OfEHTaP37xCXn67EgJV6P0hbNUkFDx8ZkTDvoDbc9q74j7AiXnuK5cDgNQoW4gLmXnRgYls0gbQW4kvwBZX0Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7078ae15a21b8dd5745ee3bd98a4a15af6ce99e4da1bbd6cd57ed131425bbd4","last_reissued_at":"2026-06-05T01:14:31.301593Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:14:31.301593Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2511.21667","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-05T01:14:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0s+s67ljhP5kwS9i68CyjOtGvgOJ0VNRq4Pz3zl1eu32dR54m6piSACYv2SHBk00aaJA9tFudOpE2IReWtb4Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-27T20:27:04.166674Z"},"content_sha256":"618d9ef182d5f6ff8dcd49d16c46333e846e3b0944c7d6555fd403d7ddcf6fc7","schema_version":"1.0","event_id":"sha256:618d9ef182d5f6ff8dcd49d16c46333e846e3b0944c7d6555fd403d7ddcf6fc7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:24DYVYK2EG4N2V2F5Y55TCSKCW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Escaping the Verifier: Learning to Reason via Demonstrations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Ivan Provilkov, Locke Cai, Max Ryabinin","submitted_at":"2025-11-26T18:42:52Z","abstract_excerpt":"Training Large Language Models (LLMs) to reason often relies on Reinforcement Learning (RL) with task-specific verifiers. However, many real-world reasoning-intensive tasks lack verifiers, despite offering abundant expert demonstrations that remain under-utilized for reasoning-focused training. We introduce RARO (Relativistic Adversarial Reasoning Optimization), which learns strong reasoning capabilities from expert demonstrations alone via Inverse Reinforcement Learning. RARO sets up an adversarial game between a policy and a relativistic critic: the policy learns to mimic expert answers, whi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.21667","kind":"arxiv","version":4},"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/2511.21667/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":"2026-06-05T01:14:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9K3WbDDXGBwSimQfHJArVsboZE1cc10gDUwdjWW/5em1geNIA0sc76urw/ue94Gw5QX06zrZ709p6IqVfVktCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-27T20:27:04.167041Z"},"content_sha256":"a5b22013c81f5304e4a9f8880749728b1d3af8bfac22a8827e25c9049e0d506c","schema_version":"1.0","event_id":"sha256:a5b22013c81f5304e4a9f8880749728b1d3af8bfac22a8827e25c9049e0d506c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/24DYVYK2EG4N2V2F5Y55TCSKCW/bundle.json","state_url":"https://pith.science/pith/24DYVYK2EG4N2V2F5Y55TCSKCW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/24DYVYK2EG4N2V2F5Y55TCSKCW/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":"2026-06-27T20:27:04Z","links":{"resolver":"https://pith.science/pith/24DYVYK2EG4N2V2F5Y55TCSKCW","bundle":"https://pith.science/pith/24DYVYK2EG4N2V2F5Y55TCSKCW/bundle.json","state":"https://pith.science/pith/24DYVYK2EG4N2V2F5Y55TCSKCW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/24DYVYK2EG4N2V2F5Y55TCSKCW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:24DYVYK2EG4N2V2F5Y55TCSKCW","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":"02bb3829985d2c84f4c6359ecc4d8f148409c51f23f53a28db65aeae38b51dbc","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-26T18:42:52Z","title_canon_sha256":"be28bb4727885b6009197b0f861d9872d1dad43c25333ac91ceb7f0a93f27334"},"schema_version":"1.0","source":{"id":"2511.21667","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2511.21667","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"arxiv_version","alias_value":"2511.21667v4","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.21667","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"pith_short_12","alias_value":"24DYVYK2EG4N","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"pith_short_16","alias_value":"24DYVYK2EG4N2V2F","created_at":"2026-06-05T01:14:31Z"},{"alias_kind":"pith_short_8","alias_value":"24DYVYK2","created_at":"2026-06-05T01:14:31Z"}],"graph_snapshots":[{"event_id":"sha256:a5b22013c81f5304e4a9f8880749728b1d3af8bfac22a8827e25c9049e0d506c","target":"graph","created_at":"2026-06-05T01:14:31Z","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/2511.21667/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Training Large Language Models (LLMs) to reason often relies on Reinforcement Learning (RL) with task-specific verifiers. However, many real-world reasoning-intensive tasks lack verifiers, despite offering abundant expert demonstrations that remain under-utilized for reasoning-focused training. We introduce RARO (Relativistic Adversarial Reasoning Optimization), which learns strong reasoning capabilities from expert demonstrations alone via Inverse Reinforcement Learning. RARO sets up an adversarial game between a policy and a relativistic critic: the policy learns to mimic expert answers, whi","authors_text":"Ivan Provilkov, Locke Cai, Max Ryabinin","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-26T18:42:52Z","title":"Escaping the Verifier: Learning to Reason via Demonstrations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.21667","kind":"arxiv","version":4},"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":"sha256:618d9ef182d5f6ff8dcd49d16c46333e846e3b0944c7d6555fd403d7ddcf6fc7","target":"record","created_at":"2026-06-05T01:14:31Z","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":"02bb3829985d2c84f4c6359ecc4d8f148409c51f23f53a28db65aeae38b51dbc","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-26T18:42:52Z","title_canon_sha256":"be28bb4727885b6009197b0f861d9872d1dad43c25333ac91ceb7f0a93f27334"},"schema_version":"1.0","source":{"id":"2511.21667","kind":"arxiv","version":4}},"canonical_sha256":"d7078ae15a21b8dd5745ee3bd98a4a15af6ce99e4da1bbd6cd57ed131425bbd4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d7078ae15a21b8dd5745ee3bd98a4a15af6ce99e4da1bbd6cd57ed131425bbd4","first_computed_at":"2026-06-05T01:14:31.301593Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-05T01:14:31.301593Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"OfEHTaP37xCXn67EgJV6P0hbNUkFDx8ZkTDvoDbc9q74j7AiXnuK5cDgNQoW4gLmXnRgYls0gbQW4kvwBZX0Cw==","signature_status":"signed_v1","signed_at":"2026-06-05T01:14:31.302389Z","signed_message":"canonical_sha256_bytes"},"source_id":"2511.21667","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:618d9ef182d5f6ff8dcd49d16c46333e846e3b0944c7d6555fd403d7ddcf6fc7","sha256:a5b22013c81f5304e4a9f8880749728b1d3af8bfac22a8827e25c9049e0d506c"],"state_sha256":"07b681b1f27195817653a3350e7a1f8761ec9e46fcd0f63f3391c6783e136dac"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DhrFyzhbzIJxxlDum/zx7TXchbwcOZIcWP003no8+Z+FVeits3bcxqkCkhkPMzXt8y7DTzi4GKxjbxWp/IUfDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-27T20:27:04.169116Z","bundle_sha256":"8159fcd68674da395ede55613eb0cdeb9a9f382a08f10c3b313caeb8c74e6275"}}