{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FY7XKCPDIB2RFDHSMQO75EI2PF","short_pith_number":"pith:FY7XKCPD","schema_version":"1.0","canonical_sha256":"2e3f7509e34075128cf2641dfe911a7969e38e3d397a8d6b2af24b7a6bd55575","source":{"kind":"arxiv","id":"2604.18701","version":3},"attestation_state":"computed","paper":{"title":"Curiosity-Critic: Cumulative Prediction Error Improvement as a Tractable Intrinsic Reward for World Model Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Curiosity-Critic uses cumulative prediction error improvement as an intrinsic reward for world model training, estimated via a co-trained critic.","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Haicheng Wang, Vin Bhaskara","submitted_at":"2026-04-20T18:01:15Z","abstract_excerpt":"Local prediction-error-based curiosity rewards focus on the current transition without considering the world model's cumulative prediction error across all visited transitions. We introduce Curiosity-Critic, which grounds its intrinsic reward in the improvement of this cumulative objective, and show that it admits a tractable per-step surrogate: the difference between the current prediction error and the asymptotic error baseline of the current state transition. We estimate this error baseline online with a learned critic co-trained alongside the world model; since the critic only has to learn"},"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":"2604.18701","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-04-20T18:01:15Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"6168d3d9a9a87b910e40adb8d3458536f23bc9972783af382d101ed5f2e0bb99","abstract_canon_sha256":"02cb1f8bece5d1793abd1b5ac478bf11832fb7f3098d349cc55835e5e8912221"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:09:58.265021Z","signature_b64":"YLqFHlXyegHtFsSHLHOTqGTvyRGqrxMccLxqbT1m93wHVdw+iG8+yHSAmPE9WXWvNstU3ouP8wE5Txh38suTBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2e3f7509e34075128cf2641dfe911a7969e38e3d397a8d6b2af24b7a6bd55575","last_reissued_at":"2026-06-19T16:09:58.264685Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:09:58.264685Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Curiosity-Critic: Cumulative Prediction Error Improvement as a Tractable Intrinsic Reward for World Model Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Curiosity-Critic uses cumulative prediction error improvement as an intrinsic reward for world model training, estimated via a co-trained critic.","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Haicheng Wang, Vin Bhaskara","submitted_at":"2026-04-20T18:01:15Z","abstract_excerpt":"Local prediction-error-based curiosity rewards focus on the current transition without considering the world model's cumulative prediction error across all visited transitions. We introduce Curiosity-Critic, which grounds its intrinsic reward in the improvement of this cumulative objective, and show that it admits a tractable per-step surrogate: the difference between the current prediction error and the asymptotic error baseline of the current state transition. We estimate this error baseline online with a learned critic co-trained alongside the world model; since the critic only has to learn"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experiments on a stochastic grid world show that Curiosity-Critic outperforms prediction-error, visitation-count, and Random Network Distillation methods in training speed and final world model accuracy.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The learned critic converges well before the world model saturates, providing a reliable online estimate of the asymptotic error baseline without oracle knowledge of the noise floor.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Curiosity-Critic rewards the improvement in cumulative prediction error via a tractable per-step surrogate (current error minus learned asymptotic baseline), outperforming prior curiosity methods in a stochastic grid world.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Curiosity-Critic uses cumulative prediction error improvement as an intrinsic reward for world model training, estimated via a co-trained critic.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a78d93f27f069f70ac4fb48d564842099c6e64772ff643b43fabf7aabf99a32c"},"source":{"id":"2604.18701","kind":"arxiv","version":3},"verdict":{"id":"f2f8dde8-a915-4511-a4bd-067560a536f7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T04:37:24.959327Z","strongest_claim":"Experiments on a stochastic grid world show that Curiosity-Critic outperforms prediction-error, visitation-count, and Random Network Distillation methods in training speed and final world model accuracy.","one_line_summary":"Curiosity-Critic rewards the improvement in cumulative prediction error via a tractable per-step surrogate (current error minus learned asymptotic baseline), outperforming prior curiosity methods in a stochastic grid world.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The learned critic converges well before the world model saturates, providing a reliable online estimate of the asymptotic error baseline without oracle knowledge of the noise floor.","pith_extraction_headline":"Curiosity-Critic uses cumulative prediction error improvement as an intrinsic reward for world model training, estimated via a co-trained critic."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.18701/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_compliance","ran_at":"2026-05-20T03:47:41.060560Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"b391193f5641f9d7723701897da469e69e25d531dc709ff9f77a287c0f5c8b96"},"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":"2604.18701","created_at":"2026-06-19T16:09:58.264736+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.18701v3","created_at":"2026-06-19T16:09:58.264736+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.18701","created_at":"2026-06-19T16:09:58.264736+00:00"},{"alias_kind":"pith_short_12","alias_value":"FY7XKCPDIB2R","created_at":"2026-06-19T16:09:58.264736+00:00"},{"alias_kind":"pith_short_16","alias_value":"FY7XKCPDIB2RFDHS","created_at":"2026-06-19T16:09:58.264736+00:00"},{"alias_kind":"pith_short_8","alias_value":"FY7XKCPD","created_at":"2026-06-19T16:09:58.264736+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.18963","citing_title":"Online Reward-Punishment Learning from Fixed-Channel Perceptual Event Streams without Environment Rewards","ref_index":8,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF","json":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF.json","graph_json":"https://pith.science/api/pith-number/FY7XKCPDIB2RFDHSMQO75EI2PF/graph.json","events_json":"https://pith.science/api/pith-number/FY7XKCPDIB2RFDHSMQO75EI2PF/events.json","paper":"https://pith.science/paper/FY7XKCPD"},"agent_actions":{"view_html":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF","download_json":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF.json","view_paper":"https://pith.science/paper/FY7XKCPD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.18701&json=true","fetch_graph":"https://pith.science/api/pith-number/FY7XKCPDIB2RFDHSMQO75EI2PF/graph.json","fetch_events":"https://pith.science/api/pith-number/FY7XKCPDIB2RFDHSMQO75EI2PF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF/action/storage_attestation","attest_author":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF/action/author_attestation","sign_citation":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF/action/citation_signature","submit_replication":"https://pith.science/pith/FY7XKCPDIB2RFDHSMQO75EI2PF/action/replication_record"}},"created_at":"2026-06-19T16:09:58.264736+00:00","updated_at":"2026-06-19T16:09:58.264736+00:00"}