{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:GDI64H2ASOLEVCX32BMVPM23EB","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":"5bde62e90ba08e809701597ea224f4219739e38baecbbf8ce3e4b49c1f13d497","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T05:46:29Z","title_canon_sha256":"d87452eb6514a3c74ed0d9eb5d6ed03a7e72653d7cecd600adad4f3300cf5a2f"},"schema_version":"1.0","source":{"id":"2605.13037","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13037","created_at":"2026-05-18T03:08:59Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13037v1","created_at":"2026-05-18T03:08:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13037","created_at":"2026-05-18T03:08:59Z"},{"alias_kind":"pith_short_12","alias_value":"GDI64H2ASOLE","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"GDI64H2ASOLEVCX3","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"GDI64H2A","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:7c947cdaadde845530269e748233c6046d9019506ba9fb9aeba34742674a228c","target":"graph","created_at":"2026-05-18T03:08:59Z","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":"On ARC-AGI-3, MAP enables frontier models to surpass near-zero baseline performance in 22 of 25 game environments. We further introduce MAP-2K, a dataset of map-then-act trajectories, and show that training on it outperforms expert execution traces."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That global exploration can efficiently acquire accurate environment-general priors and that the resulting structured cognitive map will remain valid and useful during subsequent task execution without introducing new errors or excessive overhead."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"MAP improves LLM agent reasoning by constructing a structured cognitive map of the environment before task execution, yielding performance gains on benchmarks like ARC-AGI-3 and superior training data via the new MAP-2K dataset."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"The Map-then-Act Paradigm lets LLM agents build environment maps before execution to escape trial-and-error cycles."}],"snapshot_sha256":"c0a064e91ec1dd89b35d3a987f4549b8818931849f01bd388a14f31a2e076ebe"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Current interactive LLM agents rely on goal-conditioned stepwise planning, where environmental understanding is acquired reactively during execution rather than established beforehand. This temporal inversion leads to Delayed Environmental Perception: agents must infer environmental constraints through trial-and-error, resulting in an Epistemic Bottleneck that traps them in inefficient failure cycles. Inspired by human affordance perception and cognitive map theory, we propose the Map-then-Act Paradigm (MAP), a plug-and-play framework that shifts environment understanding before execution. MAP","authors_text":"Jinwei Xiao, Lei Zhang, Mingye Zhu, Qi Gu, Xunliang Cai, Yueqing Sun, Yuxin Liu, Zhuowen Han, Ziang Ye","cross_cats":[],"headline":"The Map-then-Act Paradigm lets LLM agents build environment maps before execution to escape trial-and-error cycles.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T05:46:29Z","title":"MAP: A Map-then-Act Paradigm for Long-Horizon Interactive Agent Reasoning"},"references":{"count":50,"internal_anchors":18,"resolved_work":50,"sample":[{"cited_arxiv_id":"2303.08774","doi":"","is_internal_anchor":true,"ref_index":1,"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Self-rag: Learn- ing to retrieve, generate, and critique through self-reflection","work_id":"2070772a-8584-4ab0-99be-17b284fd4927","year":2023},{"cited_arxiv_id":"2512.10696","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Remember Me, Refine Me: A Dynamic Procedural Memory Framework for Experience-Driven Agent Evolution","work_id":"e1808e0d-921b-46dd-b221-96110080aa71","year":2025},{"cited_arxiv_id":"2506.13585","doi":"","is_internal_anchor":true,"ref_index":4,"title":"MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention","work_id":"c59fbe20-f41e-4140-a81c-40a12e7e8364","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Learning to self-verify makes language models better reasoners.CoRR, abs/2602.07594","work_id":"ff39b989-0a7a-491a-88b0-1a47366715ff","year":2026}],"snapshot_sha256":"4a9b7dfdccd1e989f5d84398ff435daf14df7eb99430c71f227a518296cc6d80"},"source":{"id":"2605.13037","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T19:43:04.508085Z","id":"42b10749-d2a2-45a5-9eca-5de9996b0616","model_set":{"reader":"grok-4.3"},"one_line_summary":"MAP improves LLM agent reasoning by constructing a structured cognitive map of the environment before task execution, yielding performance gains on benchmarks like ARC-AGI-3 and superior training data via the new MAP-2K dataset.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"The Map-then-Act Paradigm lets LLM agents build environment maps before execution to escape trial-and-error cycles.","strongest_claim":"On ARC-AGI-3, MAP enables frontier models to surpass near-zero baseline performance in 22 of 25 game environments. We further introduce MAP-2K, a dataset of map-then-act trajectories, and show that training on it outperforms expert execution traces.","weakest_assumption":"That global exploration can efficiently acquire accurate environment-general priors and that the resulting structured cognitive map will remain valid and useful during subsequent task execution without introducing new errors or excessive overhead."}},"verdict_id":"42b10749-d2a2-45a5-9eca-5de9996b0616"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:40cd851428d1c5318f989b4b4be801883db0a791877b5108be709b1e5517c8f5","target":"record","created_at":"2026-05-18T03:08:59Z","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":"5bde62e90ba08e809701597ea224f4219739e38baecbbf8ce3e4b49c1f13d497","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T05:46:29Z","title_canon_sha256":"d87452eb6514a3c74ed0d9eb5d6ed03a7e72653d7cecd600adad4f3300cf5a2f"},"schema_version":"1.0","source":{"id":"2605.13037","kind":"arxiv","version":1}},"canonical_sha256":"30d1ee1f4093964a8afbd05957b35b205aea96bbffb1c9c78d30e96b5a2940fc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"30d1ee1f4093964a8afbd05957b35b205aea96bbffb1c9c78d30e96b5a2940fc","first_computed_at":"2026-05-18T03:08:59.620200Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:08:59.620200Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZDegjItIeq8i/XOAEKPduy1kNmRprZ7ygbOlyCXdaQbNUSKO2u9Sq0bRi8jxGn6P21YaDbENGXu3/LI2eekdCA==","signature_status":"signed_v1","signed_at":"2026-05-18T03:08:59.620869Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13037","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:40cd851428d1c5318f989b4b4be801883db0a791877b5108be709b1e5517c8f5","sha256:7c947cdaadde845530269e748233c6046d9019506ba9fb9aeba34742674a228c"],"state_sha256":"df93a4dceae81f814c5befdbc99bd13768bbf1d48dd2a17f19a2286fce8c0d48"}