{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:ZFRSTE7U3VHN3TJ6MB6GLAZQLH","short_pith_number":"pith:ZFRSTE7U","schema_version":"1.0","canonical_sha256":"c9632993f4dd4eddcd3e607c65833059cb00e31332bdc84a483aef6d9922321f","source":{"kind":"arxiv","id":"2405.14751","version":2},"attestation_state":"computed","paper":{"title":"AGILE: A Novel Reinforcement Learning Framework of LLM Agents","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Guanhua Huang, Hanchong Zhang, Hang Li, Peiyuan Feng, Yichen He, Yuan Lin, Yuchen Zhang","submitted_at":"2024-05-23T16:17:44Z","abstract_excerpt":"We introduce a novel reinforcement learning framework of LLM agents named AGILE (AGent that Interacts and Learns from Environments) designed to perform complex conversational tasks with users, leveraging LLMs, memory, tools, and interactions with experts. The agent possesses capabilities beyond conversation, including reflection, tool usage, and expert consultation. We formulate the construction of such an LLM agent as a reinforcement learning (RL) problem, in which the LLM serves as the policy model. We fine-tune the LLM using labeled data of actions and the PPO algorithm. We focus on questio"},"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":"2405.14751","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-23T16:17:44Z","cross_cats_sorted":[],"title_canon_sha256":"82b1e3ac61a5141e25f83f45af1275bf5fc38d5cca9b4731926284a03a9a4f39","abstract_canon_sha256":"940a7be67f47de66dc964947af0c523e6097256821c5221840aa3919b238b0fa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:31:02.803571Z","signature_b64":"scAc4+gHIr7L7Z6Xx3uvhiliOjaoOCwIfBb3tEl5UnyNmKYqLtVFtxC4x8zifDu9HzEqDGj8AH5EQw3kllOUAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c9632993f4dd4eddcd3e607c65833059cb00e31332bdc84a483aef6d9922321f","last_reissued_at":"2026-07-05T09:31:02.803035Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:31:02.803035Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AGILE: A Novel Reinforcement Learning Framework of LLM Agents","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Guanhua Huang, Hanchong Zhang, Hang Li, Peiyuan Feng, Yichen He, Yuan Lin, Yuchen Zhang","submitted_at":"2024-05-23T16:17:44Z","abstract_excerpt":"We introduce a novel reinforcement learning framework of LLM agents named AGILE (AGent that Interacts and Learns from Environments) designed to perform complex conversational tasks with users, leveraging LLMs, memory, tools, and interactions with experts. The agent possesses capabilities beyond conversation, including reflection, tool usage, and expert consultation. We formulate the construction of such an LLM agent as a reinforcement learning (RL) problem, in which the LLM serves as the policy model. We fine-tune the LLM using labeled data of actions and the PPO algorithm. We focus on questio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.14751","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/2405.14751/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":"2405.14751","created_at":"2026-07-05T09:31:02.803095+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.14751v2","created_at":"2026-07-05T09:31:02.803095+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.14751","created_at":"2026-07-05T09:31:02.803095+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZFRSTE7U3VHN","created_at":"2026-07-05T09:31:02.803095+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZFRSTE7U3VHN3TJ6","created_at":"2026-07-05T09:31:02.803095+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZFRSTE7U","created_at":"2026-07-05T09:31:02.803095+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.03610","citing_title":"Orak: A Foundational Benchmark for Training and Evaluating LLM Agents on Diverse Video Games","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH","json":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH.json","graph_json":"https://pith.science/api/pith-number/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/graph.json","events_json":"https://pith.science/api/pith-number/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/events.json","paper":"https://pith.science/paper/ZFRSTE7U"},"agent_actions":{"view_html":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH","download_json":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH.json","view_paper":"https://pith.science/paper/ZFRSTE7U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.14751&json=true","fetch_graph":"https://pith.science/api/pith-number/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/graph.json","fetch_events":"https://pith.science/api/pith-number/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/action/storage_attestation","attest_author":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/action/author_attestation","sign_citation":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/action/citation_signature","submit_replication":"https://pith.science/pith/ZFRSTE7U3VHN3TJ6MB6GLAZQLH/action/replication_record"}},"created_at":"2026-07-05T09:31:02.803095+00:00","updated_at":"2026-07-05T09:31:02.803095+00:00"}