{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:CS5ZHYPLNFHSBUBRJE7M7Q27GR","short_pith_number":"pith:CS5ZHYPL","schema_version":"1.0","canonical_sha256":"14bb93e1eb694f20d031493ecfc35f346a183e5a8a7b146c7d246e982021c8e9","source":{"kind":"arxiv","id":"2410.14660","version":1},"attestation_state":"computed","paper":{"title":"A Large Language Model-Driven Reward Design Framework via Dynamic Feedback for Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jiafei Lyu, Jing-Wen Yang, Liangpeng Zhang, Runze Liu, Shengjie Sun, Xiu Li","submitted_at":"2024-10-18T17:51:51Z","abstract_excerpt":"Large Language Models (LLMs) have shown significant potential in designing reward functions for Reinforcement Learning (RL) tasks. However, obtaining high-quality reward code often involves human intervention, numerous LLM queries, or repetitive RL training. To address these issues, we propose CARD, a LLM-driven Reward Design framework that iteratively generates and improves reward function code. Specifically, CARD includes a Coder that generates and verifies the code, while a Evaluator provides dynamic feedback to guide the Coder in improving the code, eliminating the need for human feedback."},"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":"2410.14660","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-18T17:51:51Z","cross_cats_sorted":[],"title_canon_sha256":"15ab21bcb7c8d557e46a39a7d7906a6f959485dbba15f885894f7752be53eee7","abstract_canon_sha256":"1372cfed2315a05f80ed66cdff915e4910a8eb049f6afe955e0ecb9bb293ed3f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:22:35.023292Z","signature_b64":"uLrz4xugbmfcFpYOCR8zP76A1vKq0zAs8BoGORoeHSaKm7DdfSbTb3hzqK3OxiN26GAeO+lBdNdzXQv6nHdTAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"14bb93e1eb694f20d031493ecfc35f346a183e5a8a7b146c7d246e982021c8e9","last_reissued_at":"2026-07-05T09:22:35.022791Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:22:35.022791Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Large Language Model-Driven Reward Design Framework via Dynamic Feedback for Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jiafei Lyu, Jing-Wen Yang, Liangpeng Zhang, Runze Liu, Shengjie Sun, Xiu Li","submitted_at":"2024-10-18T17:51:51Z","abstract_excerpt":"Large Language Models (LLMs) have shown significant potential in designing reward functions for Reinforcement Learning (RL) tasks. However, obtaining high-quality reward code often involves human intervention, numerous LLM queries, or repetitive RL training. To address these issues, we propose CARD, a LLM-driven Reward Design framework that iteratively generates and improves reward function code. Specifically, CARD includes a Coder that generates and verifies the code, while a Evaluator provides dynamic feedback to guide the Coder in improving the code, eliminating the need for human feedback."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.14660","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/2410.14660/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":"2410.14660","created_at":"2026-07-05T09:22:35.022844+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.14660v1","created_at":"2026-07-05T09:22:35.022844+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.14660","created_at":"2026-07-05T09:22:35.022844+00:00"},{"alias_kind":"pith_short_12","alias_value":"CS5ZHYPLNFHS","created_at":"2026-07-05T09:22:35.022844+00:00"},{"alias_kind":"pith_short_16","alias_value":"CS5ZHYPLNFHSBUBR","created_at":"2026-07-05T09:22:35.022844+00:00"},{"alias_kind":"pith_short_8","alias_value":"CS5ZHYPL","created_at":"2026-07-05T09:22:35.022844+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/CS5ZHYPLNFHSBUBRJE7M7Q27GR","json":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR.json","graph_json":"https://pith.science/api/pith-number/CS5ZHYPLNFHSBUBRJE7M7Q27GR/graph.json","events_json":"https://pith.science/api/pith-number/CS5ZHYPLNFHSBUBRJE7M7Q27GR/events.json","paper":"https://pith.science/paper/CS5ZHYPL"},"agent_actions":{"view_html":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR","download_json":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR.json","view_paper":"https://pith.science/paper/CS5ZHYPL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.14660&json=true","fetch_graph":"https://pith.science/api/pith-number/CS5ZHYPLNFHSBUBRJE7M7Q27GR/graph.json","fetch_events":"https://pith.science/api/pith-number/CS5ZHYPLNFHSBUBRJE7M7Q27GR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR/action/storage_attestation","attest_author":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR/action/author_attestation","sign_citation":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR/action/citation_signature","submit_replication":"https://pith.science/pith/CS5ZHYPLNFHSBUBRJE7M7Q27GR/action/replication_record"}},"created_at":"2026-07-05T09:22:35.022844+00:00","updated_at":"2026-07-05T09:22:35.022844+00:00"}