{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JOHOYAFERAMIVTPWRH3Z3ITPSZ","short_pith_number":"pith:JOHOYAFE","schema_version":"1.0","canonical_sha256":"4b8eec00a488188acdf689f79da26f96597b357d676fa774e4d82472ae60ee57","source":{"kind":"arxiv","id":"2508.03680","version":1},"attestation_state":"computed","paper":{"title":"Agent Lightning: Train ANY AI Agents with Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Dongsheng Li, Luna K. Qiu, Siyun Zhao, Xufang Luo, Yuge Zhang, Yuqing Yang, Zhiyuan He, Zilong Wang","submitted_at":"2025-08-05T17:50:13Z","abstract_excerpt":"We present Agent Lightning, a flexible and extensible framework that enables Reinforcement Learning (RL)-based training of Large Language Models (LLMs) for any AI agent. Unlike existing methods that tightly couple RL training with agent or rely on sequence concatenation with masking, Agent Lightning achieves complete decoupling between agent execution and training, allowing seamless integration with existing agents developed via diverse ways (e.g., using frameworks like LangChain, OpenAI Agents SDK, AutoGen, and building from scratch) with almost ZERO code modifications. By formulating agent e"},"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":"2508.03680","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-05T17:50:13Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"9ec465f97a8d928af0585482422cd3699aac94a38abade17a37fd64c577cc648","abstract_canon_sha256":"a6665a995fbe8ae9b4ae66059652e9f06fc98db8af08ae2ffbe171e817126907"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:49:01.057308Z","signature_b64":"PGD7sk+IhldWnmLgHq+P8l2FcYeKZFOBzQH+nKiCieyvpFKNmBRnWsklGSa+R1/60kRMpMBXc4YJ+23/LhCJCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4b8eec00a488188acdf689f79da26f96597b357d676fa774e4d82472ae60ee57","last_reissued_at":"2026-07-05T11:49:01.056792Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:49:01.056792Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Agent Lightning: Train ANY AI Agents with Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Dongsheng Li, Luna K. Qiu, Siyun Zhao, Xufang Luo, Yuge Zhang, Yuqing Yang, Zhiyuan He, Zilong Wang","submitted_at":"2025-08-05T17:50:13Z","abstract_excerpt":"We present Agent Lightning, a flexible and extensible framework that enables Reinforcement Learning (RL)-based training of Large Language Models (LLMs) for any AI agent. Unlike existing methods that tightly couple RL training with agent or rely on sequence concatenation with masking, Agent Lightning achieves complete decoupling between agent execution and training, allowing seamless integration with existing agents developed via diverse ways (e.g., using frameworks like LangChain, OpenAI Agents SDK, AutoGen, and building from scratch) with almost ZERO code modifications. By formulating agent e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.03680","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/2508.03680/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":"2508.03680","created_at":"2026-07-05T11:49:01.056858+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.03680v1","created_at":"2026-07-05T11:49:01.056858+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.03680","created_at":"2026-07-05T11:49:01.056858+00:00"},{"alias_kind":"pith_short_12","alias_value":"JOHOYAFERAMI","created_at":"2026-07-05T11:49:01.056858+00:00"},{"alias_kind":"pith_short_16","alias_value":"JOHOYAFERAMIVTPW","created_at":"2026-07-05T11:49:01.056858+00:00"},{"alias_kind":"pith_short_8","alias_value":"JOHOYAFE","created_at":"2026-07-05T11:49:01.056858+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":20,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.18401","citing_title":"StepPO: Step-Aligned Policy Optimization for Agentic Reinforcement Learning","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2606.25451","citing_title":"Learning with a Single Rollout via Monte Carlo Pass@k Critic","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25161","citing_title":"TRUSTMEM: Learning Trustworthy Memory Consolidation for LLM Agents with Long-Term Memory","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22995","citing_title":"Group-Graph Policy Optimization for Long-Horizon Agentic Reinforcement Learning","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09138","citing_title":"Claw-R1: A Step-Level Data Middleware System for Agentic Reinforcement Learning","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05874","citing_title":"Evaluating Stochastic Collapse and Implicit Bias in Multimodal Large Language Models","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04484","citing_title":"AgentJet: A Flexible Swarm Training Framework for Agentic Reinforcement Learning","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22781","citing_title":"DeltaBox: Scaling Stateful AI Agents with Millisecond-Level Sandbox Checkpoint/Rollback","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22781","citing_title":"DeltaBox: Scaling Stateful AI Agents with Millisecond-Level Sandbox Checkpoint/Rollback","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2511.00413","citing_title":"Tree Training: Accelerating Agentic LLMs Training via Shared Prefix Reuse","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2603.24935","citing_title":"SABER: A Stealthy Agentic Black-Box Attack Framework for Vision-Language-Action Models","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27840","citing_title":"CastFlow: Learning Role-Specialized Agentic Workflows for Time Series Forecasting","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06642","citing_title":"StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06230","citing_title":"Safactory: A Scalable Agentic Infrastructure for Training Trustworthy Autonomous Intelligence","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18401","citing_title":"StepPO: Step-Aligned Policy Optimization for Agentic Reinforcement Learning","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18133","citing_title":"Multi-Agent Systems: From Classical Paradigms to Large Foundation Model-Enabled Futures","ref_index":100,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09459","citing_title":"From Reasoning to Agentic: Credit Assignment in Reinforcement Learning for Large Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06230","citing_title":"Safactory: A Scalable Agentic Infrastructure for Training Trustworthy Autonomous Intelligence","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04853","citing_title":"MemMachine: A Ground-Truth-Preserving Memory System for Personalized AI Agents","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02801","citing_title":"Reinforcement Learning for LLM-based Multi-Agent Systems through Orchestration Traces","ref_index":41,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ","json":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ.json","graph_json":"https://pith.science/api/pith-number/JOHOYAFERAMIVTPWRH3Z3ITPSZ/graph.json","events_json":"https://pith.science/api/pith-number/JOHOYAFERAMIVTPWRH3Z3ITPSZ/events.json","paper":"https://pith.science/paper/JOHOYAFE"},"agent_actions":{"view_html":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ","download_json":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ.json","view_paper":"https://pith.science/paper/JOHOYAFE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.03680&json=true","fetch_graph":"https://pith.science/api/pith-number/JOHOYAFERAMIVTPWRH3Z3ITPSZ/graph.json","fetch_events":"https://pith.science/api/pith-number/JOHOYAFERAMIVTPWRH3Z3ITPSZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ/action/storage_attestation","attest_author":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ/action/author_attestation","sign_citation":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ/action/citation_signature","submit_replication":"https://pith.science/pith/JOHOYAFERAMIVTPWRH3Z3ITPSZ/action/replication_record"}},"created_at":"2026-07-05T11:49:01.056858+00:00","updated_at":"2026-07-05T11:49:01.056858+00:00"}