{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:E3T6Q6SH7VAGMYXKLESUACUXUJ","short_pith_number":"pith:E3T6Q6SH","schema_version":"1.0","canonical_sha256":"26e7e87a47fd406662ea5925400a97a27140fc001abd3b706cd46334c019644a","source":{"kind":"arxiv","id":"2507.01823","version":1},"attestation_state":"computed","paper":{"title":"TD-MPC-Opt: Distilling Model-Based Multi-Task Reinforcement Learning Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.RO"],"primary_cat":"cs.LG","authors_text":"Dmytro Kuzmenko, Nadiya Shvai","submitted_at":"2025-07-02T15:38:49Z","abstract_excerpt":"We present a novel approach to knowledge transfer in model-based reinforcement learning, addressing the critical challenge of deploying large world models in resource-constrained environments. Our method efficiently distills a high-capacity multi-task agent (317M parameters) into a compact model (1M parameters) on the MT30 benchmark, significantly improving performance across diverse tasks. Our distilled model achieves a state-of-the-art normalized score of 28.45, surpassing the original 1M parameter model score of 18.93. This improvement demonstrates the ability of our distillation technique "},"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":"2507.01823","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-02T15:38:49Z","cross_cats_sorted":["cs.RO"],"title_canon_sha256":"a77e296d40996bcd3d79b0713d3b7fc8587c532959ebe53ae0d7a58f2809c21b","abstract_canon_sha256":"bf871f96011effdfe1e80d216d647e564b916d3c55303428897988849332eaa5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:30:55.918035Z","signature_b64":"OF/ocd0ns4M6xuCj47ilxo4I+ganXDmeDDvi/TGkvD5gS7GmyiatydCNAsUEX6lKpN6ihFUOSdaN1GoxF9uFDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"26e7e87a47fd406662ea5925400a97a27140fc001abd3b706cd46334c019644a","last_reissued_at":"2026-07-05T11:30:55.917276Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:30:55.917276Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TD-MPC-Opt: Distilling Model-Based Multi-Task Reinforcement Learning Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.RO"],"primary_cat":"cs.LG","authors_text":"Dmytro Kuzmenko, Nadiya Shvai","submitted_at":"2025-07-02T15:38:49Z","abstract_excerpt":"We present a novel approach to knowledge transfer in model-based reinforcement learning, addressing the critical challenge of deploying large world models in resource-constrained environments. Our method efficiently distills a high-capacity multi-task agent (317M parameters) into a compact model (1M parameters) on the MT30 benchmark, significantly improving performance across diverse tasks. Our distilled model achieves a state-of-the-art normalized score of 28.45, surpassing the original 1M parameter model score of 18.93. This improvement demonstrates the ability of our distillation technique "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.01823","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/2507.01823/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":"2507.01823","created_at":"2026-07-05T11:30:55.917348+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.01823v1","created_at":"2026-07-05T11:30:55.917348+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.01823","created_at":"2026-07-05T11:30:55.917348+00:00"},{"alias_kind":"pith_short_12","alias_value":"E3T6Q6SH7VAG","created_at":"2026-07-05T11:30:55.917348+00:00"},{"alias_kind":"pith_short_16","alias_value":"E3T6Q6SH7VAGMYXK","created_at":"2026-07-05T11:30:55.917348+00:00"},{"alias_kind":"pith_short_8","alias_value":"E3T6Q6SH","created_at":"2026-07-05T11:30:55.917348+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/E3T6Q6SH7VAGMYXKLESUACUXUJ","json":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ.json","graph_json":"https://pith.science/api/pith-number/E3T6Q6SH7VAGMYXKLESUACUXUJ/graph.json","events_json":"https://pith.science/api/pith-number/E3T6Q6SH7VAGMYXKLESUACUXUJ/events.json","paper":"https://pith.science/paper/E3T6Q6SH"},"agent_actions":{"view_html":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ","download_json":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ.json","view_paper":"https://pith.science/paper/E3T6Q6SH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.01823&json=true","fetch_graph":"https://pith.science/api/pith-number/E3T6Q6SH7VAGMYXKLESUACUXUJ/graph.json","fetch_events":"https://pith.science/api/pith-number/E3T6Q6SH7VAGMYXKLESUACUXUJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ/action/storage_attestation","attest_author":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ/action/author_attestation","sign_citation":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ/action/citation_signature","submit_replication":"https://pith.science/pith/E3T6Q6SH7VAGMYXKLESUACUXUJ/action/replication_record"}},"created_at":"2026-07-05T11:30:55.917348+00:00","updated_at":"2026-07-05T11:30:55.917348+00:00"}