{"paper":{"title":"ANCHOR: Abductive Network Construction with Hierarchical Orchestration for Reliable Probability Inference in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"ANCHOR builds dense hierarchical factor spaces from LLMs via iterative generation and clustering to support reliable Bayesian probability estimates.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guanran Luo, Jingqi Gao, Meihong Wang, Qingqiang Wu, Wentao Qiu, Zhongquan Jian","submitted_at":"2026-05-11T10:31:53Z","abstract_excerpt":"A central challenge in large-scale decision-making under incomplete information is estimating reliable probabilities. Recent approaches use Large Language Models (LLMs) to generate explanatory factors and coarse-grained probability estimates, which are then refined by a Na\\\"ive Bayes model over factor combinations. However, sparse factor spaces often yield ``unknown'' predictions, while expanding factors increases noise and spurious correlations, weakening conditional independence and degrading reliability. To address these limitations, we propose \\textsc{Anchor}, an aggregated Bayesian infere"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experiments show that ANCHOR markedly reduces ``unknown'' predictions and produces more reliable probability estimates than direct LLM baselines, achieving state-of-the-art performance while significantly reducing time and token overhead.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that iterative LLM generation plus clustering will reliably produce a hierarchical factor space that captures latent dependencies without introducing new noise or spurious correlations that degrade the causal Bayesian network.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ANCHOR constructs dense hierarchical factor spaces via LLM generation and clustering, then augments Naive Bayes with a causal Bayesian network to reduce unknown predictions and improve reliability of LLM-based probability estimates.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"ANCHOR builds dense hierarchical factor spaces from LLMs via iterative generation and clustering to support reliable Bayesian probability estimates.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"8e03b7baf41cd5112189190ac2845f23ac61f007097d05a2ebafd811a0e2bf4d"},"source":{"id":"2605.10328","kind":"arxiv","version":3},"verdict":{"id":"aa6242d3-231a-4d6e-9e66-a922bbba999c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T07:34:17.238471Z","strongest_claim":"Experiments show that ANCHOR markedly reduces ``unknown'' predictions and produces more reliable probability estimates than direct LLM baselines, achieving state-of-the-art performance while significantly reducing time and token overhead.","one_line_summary":"ANCHOR constructs dense hierarchical factor spaces via LLM generation and clustering, then augments Naive Bayes with a causal Bayesian network to reduce unknown predictions and improve reliability of LLM-based probability estimates.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that iterative LLM generation plus clustering will reliably produce a hierarchical factor space that captures latent dependencies without introducing new noise or spurious correlations that degrade the causal Bayesian network.","pith_extraction_headline":"ANCHOR builds dense hierarchical factor spaces from LLMs via iterative generation and clustering to support reliable Bayesian probability estimates."},"integrity":{"clean":false,"summary":{"advisory":2,"critical":1,"by_detector":{"doi_compliance":{"total":3,"advisory":2,"critical":1,"informational":0}},"informational":0},"endpoint":"/pith/2605.10328/integrity.json","findings":[{"note":"Identifier '10.1016/j.ijcce.2024.11' is syntactically valid but the DOI registry (doi.org) returned 404, and Crossref / OpenAlex / internal corpus also have no record. The cited work could not be located through any authoritative source.","detector":"doi_compliance","severity":"critical","ref_index":10,"audited_at":"2026-05-19T09:28:01.502751Z","detected_doi":"10.1016/j.ijcce.2024.11","finding_type":"unresolvable_identifier","verdict_class":"cross_source","detected_arxiv_id":null},{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.18653/v1/2024.emnlp-main.499.URL) was visible in the surrounding text but could not be confirmed against doi.org as printed.","detector":"doi_compliance","severity":"advisory","ref_index":13,"audited_at":"2026-05-19T09:28:01.502751Z","detected_doi":"10.18653/v1/2024.emnlp-main.499.URL","finding_type":"recoverable_identifier","verdict_class":"incontrovertible","detected_arxiv_id":null},{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.5555/2567709.2567725.URL) was visible in the surrounding text but could not be confirmed against doi.org as printed.","detector":"doi_compliance","severity":"advisory","ref_index":14,"audited_at":"2026-05-19T09:28:01.502751Z","detected_doi":"10.5555/2567709.2567725.URL","finding_type":"recoverable_identifier","verdict_class":"incontrovertible","detected_arxiv_id":null}],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T06:02:01.176640Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T15:35:51.175919Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T11:31:18.885170Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T09:28:01.502751Z","status":"completed","version":"1.0.0","findings_count":3}],"snapshot_sha256":"17d6ea332693556a8790839455ad4f394c0a6c8172b45ed50adc2276dca59f66"},"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"}