{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RUGDVG4HQONEICLREGCABKNKH3","short_pith_number":"pith:RUGDVG4H","schema_version":"1.0","canonical_sha256":"8d0c3a9b87839a440971218400a9aa3eedf7feab478502fa0cdf9fea187297ff","source":{"kind":"arxiv","id":"2505.22772","version":2},"attestation_state":"computed","paper":{"title":"Calibrated Value-Aware Model Learning with Probabilistic Environment Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Amir-massoud Farahmand, Anastasiia Pedan, Arash Ahmadian, Claas Voelcker, Igor Gilitschenski, Romina Abachi","submitted_at":"2025-05-28T18:40:49Z","abstract_excerpt":"The idea of value-aware model learning, that models should produce accurate value estimates, has gained prominence in model-based reinforcement learning. The MuZero loss, which penalizes a model's value function prediction compared to the ground-truth value function, has been utilized in several prominent empirical works in the literature. However, theoretical investigation into its strengths and weaknesses is limited. In this paper, we analyze the family of value-aware model learning losses, which includes the popular MuZero loss. We show that these losses, as normally used, are uncalibrated "},"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":"2505.22772","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-28T18:40:49Z","cross_cats_sorted":[],"title_canon_sha256":"109ce3318cbb5c75f657ac8d0c2b9bebbe271321831a5fb3567586629ef1db60","abstract_canon_sha256":"27186422fbfd705fb23509a4464c35f2c51019fc140f762dbf41d501f02152ea"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:22.576608Z","signature_b64":"jXWWnvPUvDMwVZPgFydUQfAUjTgk2nhGNlmruwIuKy/q1LLX9WgNNZIFOX4LOHtRVUPSyaJWI6hiAWfgkUAAAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8d0c3a9b87839a440971218400a9aa3eedf7feab478502fa0cdf9fea187297ff","last_reissued_at":"2026-07-05T11:18:22.576095Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:22.576095Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Calibrated Value-Aware Model Learning with Probabilistic Environment Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Amir-massoud Farahmand, Anastasiia Pedan, Arash Ahmadian, Claas Voelcker, Igor Gilitschenski, Romina Abachi","submitted_at":"2025-05-28T18:40:49Z","abstract_excerpt":"The idea of value-aware model learning, that models should produce accurate value estimates, has gained prominence in model-based reinforcement learning. The MuZero loss, which penalizes a model's value function prediction compared to the ground-truth value function, has been utilized in several prominent empirical works in the literature. However, theoretical investigation into its strengths and weaknesses is limited. In this paper, we analyze the family of value-aware model learning losses, which includes the popular MuZero loss. We show that these losses, as normally used, are uncalibrated "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.22772","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/2505.22772/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":"2505.22772","created_at":"2026-07-05T11:18:22.576154+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.22772v2","created_at":"2026-07-05T11:18:22.576154+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.22772","created_at":"2026-07-05T11:18:22.576154+00:00"},{"alias_kind":"pith_short_12","alias_value":"RUGDVG4HQONE","created_at":"2026-07-05T11:18:22.576154+00:00"},{"alias_kind":"pith_short_16","alias_value":"RUGDVG4HQONEICLR","created_at":"2026-07-05T11:18:22.576154+00:00"},{"alias_kind":"pith_short_8","alias_value":"RUGDVG4H","created_at":"2026-07-05T11:18:22.576154+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3","json":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3.json","graph_json":"https://pith.science/api/pith-number/RUGDVG4HQONEICLREGCABKNKH3/graph.json","events_json":"https://pith.science/api/pith-number/RUGDVG4HQONEICLREGCABKNKH3/events.json","paper":"https://pith.science/paper/RUGDVG4H"},"agent_actions":{"view_html":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3","download_json":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3.json","view_paper":"https://pith.science/paper/RUGDVG4H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.22772&json=true","fetch_graph":"https://pith.science/api/pith-number/RUGDVG4HQONEICLREGCABKNKH3/graph.json","fetch_events":"https://pith.science/api/pith-number/RUGDVG4HQONEICLREGCABKNKH3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3/action/storage_attestation","attest_author":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3/action/author_attestation","sign_citation":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3/action/citation_signature","submit_replication":"https://pith.science/pith/RUGDVG4HQONEICLREGCABKNKH3/action/replication_record"}},"created_at":"2026-07-05T11:18:22.576154+00:00","updated_at":"2026-07-05T11:18:22.576154+00:00"}