{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:FSM4SJL7JGJDTZTHH5WRBQZ3YN","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":"bcb89edfc5c0269c1257a96a4849fdbac48a9efca4fe11252d16014928c28beb","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:28:28Z","title_canon_sha256":"0f69912f03a536037278f3398eb9a7514173bb9f5528928c09371ee69d22f828"},"schema_version":"1.0","source":{"id":"2305.14992","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.14992","created_at":"2026-05-17T23:38:45Z"},{"alias_kind":"arxiv_version","alias_value":"2305.14992v2","created_at":"2026-05-17T23:38:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14992","created_at":"2026-05-17T23:38:45Z"},{"alias_kind":"pith_short_12","alias_value":"FSM4SJL7JGJD","created_at":"2026-05-18T12:33:33Z"},{"alias_kind":"pith_short_16","alias_value":"FSM4SJL7JGJDTZTH","created_at":"2026-05-18T12:33:33Z"},{"alias_kind":"pith_short_8","alias_value":"FSM4SJL7","created_at":"2026-05-18T12:33:33Z"}],"graph_snapshots":[{"event_id":"sha256:2e26ffd37de4c2bbc1c2adc8c1e5c6fe1ab82f3cb2a88bda7c08724cabab10e0","target":"graph","created_at":"2026-05-17T23:38:45Z","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":"RAP on LLAMA-33B surpasses CoT on GPT-4 with 33% relative improvement in a plan generation setting."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the LLM, when prompted to act as world model, produces sufficiently accurate state predictions and transition simulations to guide search without compounding errors that invalidate the planning process."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"RAP turns LLMs into dual world-model and planning agents via MCTS to generate better reasoning paths, outperforming CoT baselines and achieving 33% relative gains over GPT-4 CoT using LLaMA-33B on plan generation."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Language models can reason better by using themselves as world models and planning with tree search."}],"snapshot_sha256":"60587927e399c3a0400a8a9f486237cbcb4f99a44f4a472378d82856ff0484f2"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"65afd9f30e6abf4560f14233598d202f80622ee4c95e7e11b7b50b9f26131303"},"paper":{"abstract_excerpt":"Large language models (LLMs) have shown remarkable reasoning capabilities, especially when prompted to generate intermediate reasoning steps (e.g., Chain-of-Thought, CoT). However, LLMs can still struggle with problems that are easy for humans, such as generating action plans for executing tasks in a given environment, or performing complex math, logical, and commonsense reasoning. The deficiency stems from the key fact that LLMs lack an internal $\\textit{world model}$ to predict the world $\\textit{state}$ (e.g., environment status, intermediate variable values) and simulate long-term outcomes","authors_text":"Daisy Zhe Wang, Haodi Ma, Joshua Jiahua Hong, Shibo Hao, Yi Gu, Zhen Wang, Zhiting Hu","cross_cats":["cs.AI","cs.LG"],"headline":"Language models can reason better by using themselves as world models and planning with tree search.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:28:28Z","title":"Reasoning with Language Model is Planning with World Model"},"references":{"count":134,"internal_anchors":31,"resolved_work":134,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Alan Baddeley. 1992. Working memory. Science, 255(5044):556--559","work_id":"f36b7e9d-fb69-46ab-9df5-b0ea2ea4b066","year":1992},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Robert Eamon Briscoe. 2011. Mental imagery and the varieties of amodal perception. Pacific Philosophical Quarterly, 92(2):153--173","work_id":"6f8e00d0-b211-4677-8661-131bbfc2b45e","year":2011},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. 2020. Language models are few-shot lear","work_id":"50684699-ce18-4086-8bac-7cecd178fad0","year":2020},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Tom Bylander. 1994. The computational complexity of propositional strips planning. Artificial Intelligence, 69(1-2):165--204","work_id":"3be5a615-51c1-49c4-ab86-dff31f715a8e","year":1994},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":6,"title":"Eduardo F Camacho and Carlos Bordons Alba. 2013. Model predictive control. Springer science & business media","work_id":"775ab905-fd2a-4ddd-821e-0fc8324fe7ed","year":2013}],"snapshot_sha256":"03cd7f9af3bd37b779e1d8c431df627824a113282ccd5043b090d5cdbc1ce01f"},"source":{"id":"2305.14992","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-17T01:45:01.211709Z","id":"c1417558-31a0-4e87-9ca9-aad60b332789","model_set":{"reader":"grok-4.3"},"one_line_summary":"RAP turns LLMs into dual world-model and planning agents via MCTS to generate better reasoning paths, outperforming CoT baselines and achieving 33% relative gains over GPT-4 CoT using LLaMA-33B on plan generation.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Language models can reason better by using themselves as world models and planning with tree search.","strongest_claim":"RAP on LLAMA-33B surpasses CoT on GPT-4 with 33% relative improvement in a plan generation setting.","weakest_assumption":"That the LLM, when prompted to act as world model, produces sufficiently accurate state predictions and transition simulations to guide search without compounding errors that invalidate the planning process."}},"verdict_id":"c1417558-31a0-4e87-9ca9-aad60b332789"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:30ce3676c02ef7e085a2f1b6d7be67bacd0dfcfab9c903eab10f1cae030ae868","target":"record","created_at":"2026-05-17T23:38:45Z","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":"bcb89edfc5c0269c1257a96a4849fdbac48a9efca4fe11252d16014928c28beb","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:28:28Z","title_canon_sha256":"0f69912f03a536037278f3398eb9a7514173bb9f5528928c09371ee69d22f828"},"schema_version":"1.0","source":{"id":"2305.14992","kind":"arxiv","version":2}},"canonical_sha256":"2c99c9257f499239e6673f6d10c33bc37c8b61b0eb06d05048a5b7445fe822e0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2c99c9257f499239e6673f6d10c33bc37c8b61b0eb06d05048a5b7445fe822e0","first_computed_at":"2026-05-17T23:38:45.931891Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:45.931891Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JENwz5xbcQmAbrAYpk5AUmimyxtvbPesOYB5rcbZzvS2B4O0NWdIR33f7xxUhy0G0VbNVwfhKWMbj61eIau9AQ==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:45.932631Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.14992","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:30ce3676c02ef7e085a2f1b6d7be67bacd0dfcfab9c903eab10f1cae030ae868","sha256:2e26ffd37de4c2bbc1c2adc8c1e5c6fe1ab82f3cb2a88bda7c08724cabab10e0"],"state_sha256":"392ed5ad2795a187cf3d3e4e4b78a94053231dbc6ddda6cee03e4a62926ed5f5"}