{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:HLHTXXLSGQ7YCEBPDYY64734GK","short_pith_number":"pith:HLHTXXLS","schema_version":"1.0","canonical_sha256":"3acf3bdd72343f81102f1e31ee7f7c3294318fdc35df90b5f5ae4685304589ef","source":{"kind":"arxiv","id":"2312.10003","version":1},"attestation_state":"computed","paper":{"title":"ReST meets ReAct: Self-Improvement for Multi-Step Reasoning LLM Agent","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Daliang Li, Felix Yu, Kavya Kopparapu, Manzil Zaheer, Pranesh Srinivasan, Renat Aksitov, Ruiqi Guo, Sanjiv Kumar, Sheila Babayan, Sobhan Miryoosefi, Sushant Prakash, Zachary Fisher, Zonglin Li","submitted_at":"2023-12-15T18:20:15Z","abstract_excerpt":"Answering complex natural language questions often necessitates multi-step reasoning and integrating external information. Several systems have combined knowledge retrieval with a large language model (LLM) to answer such questions. These systems, however, suffer from various failure cases, and we cannot directly train them end-to-end to fix such failures, as interaction with external knowledge is non-differentiable. To address these deficiencies, we define a ReAct-style LLM agent with the ability to reason and act upon external knowledge. We further refine the agent through a ReST-like method"},"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":"2312.10003","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T18:20:15Z","cross_cats_sorted":[],"title_canon_sha256":"99f1a6bb2501bcf17139cdc3d3b7255f215c014a8a45a87edd48b555ce21458d","abstract_canon_sha256":"87ce31fc69270cb6014c1fe4550a7f241b98c43c97882d0b400a7d142294161e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:24:43.547801Z","signature_b64":"rPf5khXxU/4apqPTXBo2YNpW8Vwno0potTuP7ya0fKFLVZsP2M+K6WSwJojO19Azmsdq046pnSGHs+wDC7FpBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3acf3bdd72343f81102f1e31ee7f7c3294318fdc35df90b5f5ae4685304589ef","last_reissued_at":"2026-07-05T07:24:43.547319Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:24:43.547319Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReST meets ReAct: Self-Improvement for Multi-Step Reasoning LLM Agent","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Daliang Li, Felix Yu, Kavya Kopparapu, Manzil Zaheer, Pranesh Srinivasan, Renat Aksitov, Ruiqi Guo, Sanjiv Kumar, Sheila Babayan, Sobhan Miryoosefi, Sushant Prakash, Zachary Fisher, Zonglin Li","submitted_at":"2023-12-15T18:20:15Z","abstract_excerpt":"Answering complex natural language questions often necessitates multi-step reasoning and integrating external information. Several systems have combined knowledge retrieval with a large language model (LLM) to answer such questions. These systems, however, suffer from various failure cases, and we cannot directly train them end-to-end to fix such failures, as interaction with external knowledge is non-differentiable. To address these deficiencies, we define a ReAct-style LLM agent with the ability to reason and act upon external knowledge. We further refine the agent through a ReST-like method"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.10003","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/2312.10003/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":"2312.10003","created_at":"2026-07-05T07:24:43.547379+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.10003v1","created_at":"2026-07-05T07:24:43.547379+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.10003","created_at":"2026-07-05T07:24:43.547379+00:00"},{"alias_kind":"pith_short_12","alias_value":"HLHTXXLSGQ7Y","created_at":"2026-07-05T07:24:43.547379+00:00"},{"alias_kind":"pith_short_16","alias_value":"HLHTXXLSGQ7YCEBP","created_at":"2026-07-05T07:24:43.547379+00:00"},{"alias_kind":"pith_short_8","alias_value":"HLHTXXLS","created_at":"2026-07-05T07:24:43.547379+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21090","citing_title":"Self-Improvement Can Self-Regress: The Rise-and-Collapse Failure Mode of LLM Self-Training","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09105","citing_title":"Graph2Idea:Retrieval-Augmented Scientific Idea Generation with Graph-Structured Contexts","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00038","citing_title":"Stop Hand-Holding Your Coding Agent: Engineering the Loops that Replace Step-by-Step Prompting","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07074","citing_title":"SlimSearcher: Training Efficiency-Aware Web Agents via Adaptive Reward Gating","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30961","citing_title":"EvoGens: A Population-Based Heuristic Search Framework for Scientific Idea Generation","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17900","citing_title":"DuIVRS-2: An LLM-based Interactive Voice Response System for Large-scale POI Attribute Acquisition","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2601.12538","citing_title":"Agentic Reasoning for Large Language Models","ref_index":103,"is_internal_anchor":false},{"citing_arxiv_id":"2504.19678","citing_title":"From LLM Reasoning to Autonomous AI Agents: A Comprehensive Review","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2511.20857","citing_title":"Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory","ref_index":234,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03976","citing_title":"Quantifying Trust: Financial Risk Management for Trustworthy AI Agents","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK","json":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK.json","graph_json":"https://pith.science/api/pith-number/HLHTXXLSGQ7YCEBPDYY64734GK/graph.json","events_json":"https://pith.science/api/pith-number/HLHTXXLSGQ7YCEBPDYY64734GK/events.json","paper":"https://pith.science/paper/HLHTXXLS"},"agent_actions":{"view_html":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK","download_json":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK.json","view_paper":"https://pith.science/paper/HLHTXXLS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.10003&json=true","fetch_graph":"https://pith.science/api/pith-number/HLHTXXLSGQ7YCEBPDYY64734GK/graph.json","fetch_events":"https://pith.science/api/pith-number/HLHTXXLSGQ7YCEBPDYY64734GK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK/action/storage_attestation","attest_author":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK/action/author_attestation","sign_citation":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK/action/citation_signature","submit_replication":"https://pith.science/pith/HLHTXXLSGQ7YCEBPDYY64734GK/action/replication_record"}},"created_at":"2026-07-05T07:24:43.547379+00:00","updated_at":"2026-07-05T07:24:43.547379+00:00"}