{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2D4IAYJCQ6PTGFHTE5NJD66436","short_pith_number":"pith:2D4IAYJC","schema_version":"1.0","canonical_sha256":"d0f8806122879f3314f3275a91fbdcdfb275c6d4a833876681389a87a8b42cd1","source":{"kind":"arxiv","id":"2503.06580","version":1},"attestation_state":"computed","paper":{"title":"Agent models: Internalizing Chain-of-Action Generation into Reasoning models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiangming Shu, Jitao Sang, Xinyan Wen, Yuqi Yang, Yuxiang Zhang","submitted_at":"2025-03-09T12:19:47Z","abstract_excerpt":"Traditional agentic workflows rely on external prompts to manage interactions with tools and the environment, which limits the autonomy of reasoning models. We position \\emph{Large Agent Models (LAMs)} that internalize the generation of \\emph{Chain-of-Action (CoA)}, enabling the model to autonomously decide when and how to use external tools. Our proposed AutoCoA framework combines supervised fine-tuning (SFT) and reinforcement learning (RL), allowing the model to seamlessly switch between reasoning and action while efficiently managing environment interactions. Main components include step-le"},"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":"2503.06580","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-03-09T12:19:47Z","cross_cats_sorted":[],"title_canon_sha256":"ea5ea65db356117d5824bdc30647eb3c69aa21eccaf45445241d64b84e5b09aa","abstract_canon_sha256":"f133071e64155593e498c31bf0d654f0aeaafd4926069c849f61dcdb34be754e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:27:37.571024Z","signature_b64":"GQJuPC5ey+SRqa7OzajWyTBykOR/2FGuBr4dJ1k6MmaJDXMYzH0v/vtvSY22CFZ5hP/OnRLMU+XXI8PTNpusDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0f8806122879f3314f3275a91fbdcdfb275c6d4a833876681389a87a8b42cd1","last_reissued_at":"2026-07-05T10:27:37.570354Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:27:37.570354Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Agent models: Internalizing Chain-of-Action Generation into Reasoning models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiangming Shu, Jitao Sang, Xinyan Wen, Yuqi Yang, Yuxiang Zhang","submitted_at":"2025-03-09T12:19:47Z","abstract_excerpt":"Traditional agentic workflows rely on external prompts to manage interactions with tools and the environment, which limits the autonomy of reasoning models. We position \\emph{Large Agent Models (LAMs)} that internalize the generation of \\emph{Chain-of-Action (CoA)}, enabling the model to autonomously decide when and how to use external tools. Our proposed AutoCoA framework combines supervised fine-tuning (SFT) and reinforcement learning (RL), allowing the model to seamlessly switch between reasoning and action while efficiently managing environment interactions. Main components include step-le"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.06580","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/2503.06580/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":"2503.06580","created_at":"2026-07-05T10:27:37.570437+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.06580v1","created_at":"2026-07-05T10:27:37.570437+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.06580","created_at":"2026-07-05T10:27:37.570437+00:00"},{"alias_kind":"pith_short_12","alias_value":"2D4IAYJCQ6PT","created_at":"2026-07-05T10:27:37.570437+00:00"},{"alias_kind":"pith_short_16","alias_value":"2D4IAYJCQ6PTGFHT","created_at":"2026-07-05T10:27:37.570437+00:00"},{"alias_kind":"pith_short_8","alias_value":"2D4IAYJC","created_at":"2026-07-05T10:27:37.570437+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.00861","citing_title":"Erase to Improve: Erasable Reinforcement Learning for Search-Augmented LLMs","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436","json":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436.json","graph_json":"https://pith.science/api/pith-number/2D4IAYJCQ6PTGFHTE5NJD66436/graph.json","events_json":"https://pith.science/api/pith-number/2D4IAYJCQ6PTGFHTE5NJD66436/events.json","paper":"https://pith.science/paper/2D4IAYJC"},"agent_actions":{"view_html":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436","download_json":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436.json","view_paper":"https://pith.science/paper/2D4IAYJC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.06580&json=true","fetch_graph":"https://pith.science/api/pith-number/2D4IAYJCQ6PTGFHTE5NJD66436/graph.json","fetch_events":"https://pith.science/api/pith-number/2D4IAYJCQ6PTGFHTE5NJD66436/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436/action/storage_attestation","attest_author":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436/action/author_attestation","sign_citation":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436/action/citation_signature","submit_replication":"https://pith.science/pith/2D4IAYJCQ6PTGFHTE5NJD66436/action/replication_record"}},"created_at":"2026-07-05T10:27:37.570437+00:00","updated_at":"2026-07-05T10:27:37.570437+00:00"}