{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6UPARSE23RA6KURLSHNKU2S7LB","short_pith_number":"pith:6UPARSE2","schema_version":"1.0","canonical_sha256":"f51e08c89adc41e5522b91daaa6a5f5861d2d79c15aa009f3680b76ba3ca8611","source":{"kind":"arxiv","id":"2505.20732","version":1},"attestation_state":"computed","paper":{"title":"SPA-RL: Reinforcing LLM Agents via Stepwise Progress Attribution","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chak Tou Leong, Hanlin Wang, Jian Wang, Jiashuo Wang, Wenjie Li","submitted_at":"2025-05-27T05:21:04Z","abstract_excerpt":"Reinforcement learning (RL) holds significant promise for training LLM agents to handle complex, goal-oriented tasks that require multi-step interactions with external environments. However, a critical challenge when applying RL to these agentic tasks arises from delayed rewards: feedback signals are typically available only after the entire task is completed. This makes it non-trivial to assign delayed rewards to earlier actions, providing insufficient guidance regarding environmental constraints and hindering agent training. In this work, we draw on the insight that the ultimate completion o"},"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":"2505.20732","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-27T05:21:04Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"dd0a571591063ba727b55b6a537d533010c9b9f9c8b89384524ad6954349edca","abstract_canon_sha256":"66624f54845599cf7d24dee28c280139de567973755ae93120d41a145787d69c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:10:17.657446Z","signature_b64":"D0xotvFEYuKlgKA3zxG3J1TE9hygUDvXdd9/5SJpWhdXHv++TGocIIGsfGZnQPfskYBrIjBr2ZoefVgH8GGnCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f51e08c89adc41e5522b91daaa6a5f5861d2d79c15aa009f3680b76ba3ca8611","last_reissued_at":"2026-07-05T11:10:17.657022Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:10:17.657022Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SPA-RL: Reinforcing LLM Agents via Stepwise Progress Attribution","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chak Tou Leong, Hanlin Wang, Jian Wang, Jiashuo Wang, Wenjie Li","submitted_at":"2025-05-27T05:21:04Z","abstract_excerpt":"Reinforcement learning (RL) holds significant promise for training LLM agents to handle complex, goal-oriented tasks that require multi-step interactions with external environments. However, a critical challenge when applying RL to these agentic tasks arises from delayed rewards: feedback signals are typically available only after the entire task is completed. This makes it non-trivial to assign delayed rewards to earlier actions, providing insufficient guidance regarding environmental constraints and hindering agent training. In this work, we draw on the insight that the ultimate completion o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.20732","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/2505.20732/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":"2505.20732","created_at":"2026-07-05T11:10:17.657080+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.20732v1","created_at":"2026-07-05T11:10:17.657080+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.20732","created_at":"2026-07-05T11:10:17.657080+00:00"},{"alias_kind":"pith_short_12","alias_value":"6UPARSE23RA6","created_at":"2026-07-05T11:10:17.657080+00:00"},{"alias_kind":"pith_short_16","alias_value":"6UPARSE23RA6KURL","created_at":"2026-07-05T11:10:17.657080+00:00"},{"alias_kind":"pith_short_8","alias_value":"6UPARSE2","created_at":"2026-07-05T11:10:17.657080+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":21,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25852","citing_title":"Semantic Consistency Policy Optimization for Reinforcement Learning of LLM Agents","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25451","citing_title":"Learning with a Single Rollout via Monte Carlo Pass@k Critic","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2606.12908","citing_title":"SENTINEL: Failure-Driven Reinforcement Learning for Training Tool-Using Language Model Agents","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10507","citing_title":"HIPIF: Hierarchical Planning and Information Folding for Long-Horizon LLM Agent Learning","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02373","citing_title":"Harness-1: Reinforcement Learning for Search Agents with State-Externalizing Harnesses","ref_index":109,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02372","citing_title":"COMAP: Co-Evolving World Models and Agent Policies for LLM Agents","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01249","citing_title":"Trust Region On-Policy Distillation","ref_index":127,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17871","citing_title":"StepGuard: Guarding Web Navigation via Single-Step Calibration","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2504.01990","citing_title":"Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems","ref_index":170,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17877","citing_title":"PAIR: Prefix-Aware Internal Reward Model for Multi-Turn Agent Optimization","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2509.02547","citing_title":"The Landscape of Agentic Reinforcement Learning for LLMs: A Survey","ref_index":128,"is_internal_anchor":false},{"citing_arxiv_id":"2507.21046","citing_title":"A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence","ref_index":239,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09934","citing_title":"TRACER: Verifiable Generative Provenance for Multimodal Tool-Using Agents","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00425","citing_title":"AEM: Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00425","citing_title":"AEM: Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06130","citing_title":"Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08013","citing_title":"Learning CLI Agents with Structured Action Credit under Selective Observation","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07462","citing_title":"The Moltbook Files: A Harmless Slopocalypse or Humanity's Last Experiment","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02572","citing_title":"On Training Large Language Models for Long-Horizon Tasks: An Empirical Study of Horizon Length","ref_index":43,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB","json":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB.json","graph_json":"https://pith.science/api/pith-number/6UPARSE23RA6KURLSHNKU2S7LB/graph.json","events_json":"https://pith.science/api/pith-number/6UPARSE23RA6KURLSHNKU2S7LB/events.json","paper":"https://pith.science/paper/6UPARSE2"},"agent_actions":{"view_html":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB","download_json":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB.json","view_paper":"https://pith.science/paper/6UPARSE2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.20732&json=true","fetch_graph":"https://pith.science/api/pith-number/6UPARSE23RA6KURLSHNKU2S7LB/graph.json","fetch_events":"https://pith.science/api/pith-number/6UPARSE23RA6KURLSHNKU2S7LB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB/action/storage_attestation","attest_author":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB/action/author_attestation","sign_citation":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB/action/citation_signature","submit_replication":"https://pith.science/pith/6UPARSE23RA6KURLSHNKU2S7LB/action/replication_record"}},"created_at":"2026-07-05T11:10:17.657080+00:00","updated_at":"2026-07-05T11:10:17.657080+00:00"}