{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XOKSYG3K3JCS2WKA2ZBDLWER4B","short_pith_number":"pith:XOKSYG3K","schema_version":"1.0","canonical_sha256":"bb952c1b6ada452d5940d64235d891e06d5c41be14d988cd43c909060c2b5423","source":{"kind":"arxiv","id":"2410.18528","version":1},"attestation_state":"computed","paper":{"title":"PRACT: Optimizing Principled Reasoning and Acting of LLM Agent","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Caiming Xiong, Huan Wang, Jianguo Zhang, Juan Carlos Niebles, Juntao Tan, Liangwei Yang, Ming Zhu, Rithesh Murthy, Shelby Heinecke, Shirley Kokane, Silvio Savarese, Thai Hoang, Tian Lan, Weiran Yao, Zhiwei Liu, Zuxin Liu","submitted_at":"2024-10-24T08:21:51Z","abstract_excerpt":"We introduce the Principled Reasoning and Acting (PRAct) framework, a novel method for learning and enforcing action principles from trajectory data. Central to our approach is the use of text gradients from a reflection and optimization engine to derive these action principles. To adapt action principles to specific task requirements, we propose a new optimization framework, Reflective Principle Optimization (RPO). After execution, RPO employs a reflector to critique current action principles and an optimizer to update them accordingly. We develop the RPO framework under two scenarios: Reward"},"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":"2410.18528","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-10-24T08:21:51Z","cross_cats_sorted":[],"title_canon_sha256":"2299c4c33621b13fdd002947cada0791682c884bca6f4f0f7a5ba4399060f8be","abstract_canon_sha256":"6c17f35f4f5dcafbfc3d75c27c4d51d33608035e6dc29c75844c1410e5ae892d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:25:09.760456Z","signature_b64":"4LpvGbtdEwOESYXpnm0C8QyKawGyN4p4QKvPDhBctJ9OcWPz+KwrXR40OGqMqrR5BHAvihXGyJwPbj4AySWBCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bb952c1b6ada452d5940d64235d891e06d5c41be14d988cd43c909060c2b5423","last_reissued_at":"2026-07-05T09:25:09.759973Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:25:09.759973Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PRACT: Optimizing Principled Reasoning and Acting of LLM Agent","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Caiming Xiong, Huan Wang, Jianguo Zhang, Juan Carlos Niebles, Juntao Tan, Liangwei Yang, Ming Zhu, Rithesh Murthy, Shelby Heinecke, Shirley Kokane, Silvio Savarese, Thai Hoang, Tian Lan, Weiran Yao, Zhiwei Liu, Zuxin Liu","submitted_at":"2024-10-24T08:21:51Z","abstract_excerpt":"We introduce the Principled Reasoning and Acting (PRAct) framework, a novel method for learning and enforcing action principles from trajectory data. Central to our approach is the use of text gradients from a reflection and optimization engine to derive these action principles. To adapt action principles to specific task requirements, we propose a new optimization framework, Reflective Principle Optimization (RPO). After execution, RPO employs a reflector to critique current action principles and an optimizer to update them accordingly. We develop the RPO framework under two scenarios: Reward"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.18528","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/2410.18528/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":"2410.18528","created_at":"2026-07-05T09:25:09.760032+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.18528v1","created_at":"2026-07-05T09:25:09.760032+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.18528","created_at":"2026-07-05T09:25:09.760032+00:00"},{"alias_kind":"pith_short_12","alias_value":"XOKSYG3K3JCS","created_at":"2026-07-05T09:25:09.760032+00:00"},{"alias_kind":"pith_short_16","alias_value":"XOKSYG3K3JCS2WKA","created_at":"2026-07-05T09:25:09.760032+00:00"},{"alias_kind":"pith_short_8","alias_value":"XOKSYG3K","created_at":"2026-07-05T09:25:09.760032+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/XOKSYG3K3JCS2WKA2ZBDLWER4B","json":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B.json","graph_json":"https://pith.science/api/pith-number/XOKSYG3K3JCS2WKA2ZBDLWER4B/graph.json","events_json":"https://pith.science/api/pith-number/XOKSYG3K3JCS2WKA2ZBDLWER4B/events.json","paper":"https://pith.science/paper/XOKSYG3K"},"agent_actions":{"view_html":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B","download_json":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B.json","view_paper":"https://pith.science/paper/XOKSYG3K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.18528&json=true","fetch_graph":"https://pith.science/api/pith-number/XOKSYG3K3JCS2WKA2ZBDLWER4B/graph.json","fetch_events":"https://pith.science/api/pith-number/XOKSYG3K3JCS2WKA2ZBDLWER4B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B/action/storage_attestation","attest_author":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B/action/author_attestation","sign_citation":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B/action/citation_signature","submit_replication":"https://pith.science/pith/XOKSYG3K3JCS2WKA2ZBDLWER4B/action/replication_record"}},"created_at":"2026-07-05T09:25:09.760032+00:00","updated_at":"2026-07-05T09:25:09.760032+00:00"}