{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:2WVDHOSVEUUYFOMW5F73ZXKO36","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":"2249ea6f2d705073f5114f9884c3e0f72bec7de4981d8fcd62e9e20c1a5de161","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-03-22T14:14:23Z","title_canon_sha256":"361560d9f934dff808203e8a595b79d66b15d50e761f70dd811e5c9fa163ea41"},"schema_version":"1.0","source":{"id":"2603.21250","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.21250","created_at":"2026-05-17T23:39:15Z"},{"alias_kind":"arxiv_version","alias_value":"2603.21250v2","created_at":"2026-05-17T23:39:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.21250","created_at":"2026-05-17T23:39:15Z"},{"alias_kind":"pith_short_12","alias_value":"2WVDHOSVEUUY","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"2WVDHOSVEUUYFOMW","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"2WVDHOSV","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:4d690b105be8d09eb495a7e25d7e5bba67072f97bd47de9c369421283e4f1122","target":"graph","created_at":"2026-05-17T23:39:15Z","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":"GoS significantly outperforms all baselines, providing a robust solution for complex abductive tasks."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That grounding multi-agent LLM collaboration in a causal graph and state machine will reliably prevent Evidence Fabrication, Context Drift, Failed Backtracking, and Early Stopping without introducing new failure modes."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"GoS structures LLM reasoning with belief-state graphs and state machines to enable reliable abductive inference, outperforming baselines on two real-world datasets."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Graph of States uses a causal graph and state machine to guide LLMs through reliable abductive reasoning."}],"snapshot_sha256":"bd9079580b9fce65634dcde1fc5fef0a0b667cf4e3fab6c028eaaa8a1c562de8"},"formal_canon":{"evidence_count":1,"snapshot_sha256":"ce4d8074235231b1cca4ed49c074cd6160a2b69528e2d5b0ace575bdf8869bc1"},"paper":{"abstract_excerpt":"Logical reasoning encompasses deduction, induction, and abduction. However, while Large Language Models (LLMs) have effectively mastered the former two, abductive reasoning remains significantly underexplored. Existing frameworks, predominantly designed for static deductive tasks, fail to generalize to abductive reasoning due to unstructured state representation and lack of explicit state control. Consequently, they are inevitably prone to Evidence Fabrication, Context Drift, Failed Backtracking, and Early Stopping. To bridge this gap, we introduce Graph of States (GoS), a general-purpose neur","authors_text":"Dan Pei, Jiamin Jiang, Jiasong Feng, Lu Teng, Qingliang Zhang, Rongchen Gao, Shenglin Zhang, Tong Liu, Wenjie Zhang, Xidao Wen, Yongqian Sun, Yu Luo","cross_cats":[],"headline":"Graph of States uses a causal graph and state machine to guide LLMs through reliable abductive reasoning.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-03-22T14:14:23Z","title":"Graph of States: Solving Abductive Tasks with Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.21250","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-15T07:16:37.521311Z","id":"c6b40fc3-a590-4601-b28c-98f9bbea01f9","model_set":{"reader":"grok-4.3"},"one_line_summary":"GoS structures LLM reasoning with belief-state graphs and state machines to enable reliable abductive inference, outperforming baselines on two real-world datasets.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Graph of States uses a causal graph and state machine to guide LLMs through reliable abductive reasoning.","strongest_claim":"GoS significantly outperforms all baselines, providing a robust solution for complex abductive tasks.","weakest_assumption":"That grounding multi-agent LLM collaboration in a causal graph and state machine will reliably prevent Evidence Fabrication, Context Drift, Failed Backtracking, and Early Stopping without introducing new failure modes."}},"verdict_id":"c6b40fc3-a590-4601-b28c-98f9bbea01f9"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:1a0c822f81f5236e5b45871b127082f7ad2efe16d3b3049437918f6b3929e62a","target":"record","created_at":"2026-05-17T23:39:15Z","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":"2249ea6f2d705073f5114f9884c3e0f72bec7de4981d8fcd62e9e20c1a5de161","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-03-22T14:14:23Z","title_canon_sha256":"361560d9f934dff808203e8a595b79d66b15d50e761f70dd811e5c9fa163ea41"},"schema_version":"1.0","source":{"id":"2603.21250","kind":"arxiv","version":2}},"canonical_sha256":"d5aa33ba55252982b996e97fbcdd4edf91f030e93bdb9f8d690c777929feca8b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d5aa33ba55252982b996e97fbcdd4edf91f030e93bdb9f8d690c777929feca8b","first_computed_at":"2026-05-17T23:39:15.705113Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:15.705113Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rkXpFBZBsl1zVQN4Y39gKFW5ZcZYcKLjeKHkcl+1wQ6ruOxU8/EmZL3q0PwjI/ZX/wGm3i2ZIV9xp0Z6CM67Dw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:15.705631Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.21250","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1a0c822f81f5236e5b45871b127082f7ad2efe16d3b3049437918f6b3929e62a","sha256:4d690b105be8d09eb495a7e25d7e5bba67072f97bd47de9c369421283e4f1122"],"state_sha256":"b1a413d44b8da0cd346858d96879f3178799b298941a49dfa6f67cd12319ad42"}