{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:GBMGNK6Q6ZHYSDHGILMEQ2UDMC","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"d92a6bb55826794aaea03e8bdd224d72c737bc0a610264b10ccbaf710c99a3c9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-20T15:22:39Z","title_canon_sha256":"9b0768938686a95fe0912f58ee7dfc52476fb36ae3bc01c12c7a51db13f1f769"},"schema_version":"1.0","source":{"id":"2604.18401","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.18401","created_at":"2026-06-02T02:04:53Z"},{"alias_kind":"arxiv_version","alias_value":"2604.18401v2","created_at":"2026-06-02T02:04:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.18401","created_at":"2026-06-02T02:04:53Z"},{"alias_kind":"pith_short_12","alias_value":"GBMGNK6Q6ZHY","created_at":"2026-06-02T02:04:53Z"},{"alias_kind":"pith_short_16","alias_value":"GBMGNK6Q6ZHYSDHG","created_at":"2026-06-02T02:04:53Z"},{"alias_kind":"pith_short_8","alias_value":"GBMGNK6Q","created_at":"2026-06-02T02:04:53Z"}],"graph_snapshots":[{"event_id":"sha256:faa5bea6619c1b7be7b6effa563fffc9c034b59b1a8207afb7a59289f5fc5519","target":"graph","created_at":"2026-06-02T02:04:53Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"the conventional token-level Markov Decision Process (MDP) should be advanced to a step-level MDP formulation, and that the step, rather than the token, should be regarded as the proper action representation for LLM agents"},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That redefining the MDP and credit assignment at step granularity will meaningfully address delayed/sparse rewards and long context challenges in multi-turn agent settings."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"StepPO argues that LLM agents should optimize at the step level rather than token level to better handle delayed rewards and long contexts in agentic RL."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"LLM agents need step-level MDP and credit assignment rather than token-level modeling for multi-turn RL."}],"snapshot_sha256":"a13be7d3ed31acb1022e48c7c1388a19ea717b0a6749fc0fe81cb0f2cbaefa1d"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-20T04:02:34.505902Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2604.18401/integrity.json","findings":[],"snapshot_sha256":"e916b5c1865fb0d190821c203dfeac2c3e81eb7a3f5b3bc71143c18d8b737b04","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Agentic reinforcement learning (RL) is emerging as a critical post-training paradigm for improving LLM agent capabilities. Existing RL algorithms for LLMs largely follow the token-centric paradigm as in RLHF and RLVR, where tokens serve as the basic units for modeling and optimization. However, this paradigm introduces a granularity mismatch in agentic RL, as it optimizes token-level predictions while LLM agents make step-level decisions through cycles of environmental observations and actions. To bridge this gap, we propose \\textbf{StepPO}, a step-centric paradigm for agentic RL via step-alig","authors_text":"Daoyu Wang, Enhong Chen, Jie Ouyang, Mingyue Cheng, Qi Liu, Qingchuan Li, Shuo Yu","cross_cats":[],"headline":"LLM agents need step-level MDP and credit assignment rather than token-level modeling for multi-turn RL.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-20T15:22:39Z","title":"StepPO: Step-Aligned Policy Optimization for Agentic Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2604.18401","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-10T04:23:57.354913Z","id":"9f7497bf-1d29-4123-a5db-df28933e1a61","model_set":{"reader":"grok-4.3"},"one_line_summary":"StepPO argues that LLM agents should optimize at the step level rather than token level to better handle delayed rewards and long contexts in agentic RL.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"LLM agents need step-level MDP and credit assignment rather than token-level modeling for multi-turn RL.","strongest_claim":"the conventional token-level Markov Decision Process (MDP) should be advanced to a step-level MDP formulation, and that the step, rather than the token, should be regarded as the proper action representation for LLM agents","weakest_assumption":"That redefining the MDP and credit assignment at step granularity will meaningfully address delayed/sparse rewards and long context challenges in multi-turn agent settings."}},"verdict_id":"9f7497bf-1d29-4123-a5db-df28933e1a61"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9f46ec284695587367947198d5a66dd591feb7e6566293249c3c1c177b5988a3","target":"record","created_at":"2026-06-02T02:04:53Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"d92a6bb55826794aaea03e8bdd224d72c737bc0a610264b10ccbaf710c99a3c9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-20T15:22:39Z","title_canon_sha256":"9b0768938686a95fe0912f58ee7dfc52476fb36ae3bc01c12c7a51db13f1f769"},"schema_version":"1.0","source":{"id":"2604.18401","kind":"arxiv","version":2}},"canonical_sha256":"305866abd0f64f890ce642d8486a8360a67d518fef2ed4d01d29bb17a1bed39f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"305866abd0f64f890ce642d8486a8360a67d518fef2ed4d01d29bb17a1bed39f","first_computed_at":"2026-06-02T02:04:53.338134Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T02:04:53.338134Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"yP2Nsh2+YOecJiM6Sgddi8eJaX1kpiPmAf3X/NE/HptA/41AYPdSz51awUFlMMzJSiI/5kdfbRtdJetnn96ZBQ==","signature_status":"signed_v1","signed_at":"2026-06-02T02:04:53.338536Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.18401","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9f46ec284695587367947198d5a66dd591feb7e6566293249c3c1c177b5988a3","sha256:faa5bea6619c1b7be7b6effa563fffc9c034b59b1a8207afb7a59289f5fc5519"],"state_sha256":"68e2814d6d781c36918ea53dc363588e4b712285a381423071e3bd65bcd01d41"}