{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QP422M6HM6FFCPS7JQPVUNLMWX","short_pith_number":"pith:QP422M6H","schema_version":"1.0","canonical_sha256":"83f9ad33c7678a513e5f4c1f5a356cb5f22bd3b3f170c3ce303adcec58612230","source":{"kind":"arxiv","id":"2504.01282","version":2},"attestation_state":"computed","paper":{"title":"Prompt-Reverse Inconsistency: LLM Self-Inconsistency Beyond Generative Randomness and Prompt Paraphrasing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jihyun Janice Ahn, Wenpeng Yin","submitted_at":"2025-04-02T01:19:37Z","abstract_excerpt":"While the inconsistency of LLMs is not a novel topic, prior research has predominantly addressed two types of generative inconsistencies: i) Randomness Inconsistency: running the same LLM multiple trials, yielding varying responses; ii) Paraphrase Inconsistency: paraphrased prompts result in different responses from the same LLM. Randomness Inconsistency arises from the inherent randomness due to stochastic sampling in generative models, while Paraphrase Inconsistency is a consequence of the language modeling objectives, where paraphrased prompts alter the distribution of vocabulary logits. Th"},"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":"2504.01282","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-02T01:19:37Z","cross_cats_sorted":[],"title_canon_sha256":"918139fd22e858e2e03d3f9181498c801f6c30a8be57f3cf6f3b583552a4089c","abstract_canon_sha256":"96fdf1d044bbc7d03cae6fd5081f7c23b51fc3c95616fe56ec6d686c04c00279"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:45:33.390270Z","signature_b64":"CZpkFlM9hFQ0bQM3qh1Rffs6msrDABlKsxWIEqi4TLPp8wQDa7xbtsG+LbEpIkQDEXQD61Zp3QYTcPz6qXukBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"83f9ad33c7678a513e5f4c1f5a356cb5f22bd3b3f170c3ce303adcec58612230","last_reissued_at":"2026-07-05T11:45:33.389718Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:45:33.389718Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Prompt-Reverse Inconsistency: LLM Self-Inconsistency Beyond Generative Randomness and Prompt Paraphrasing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jihyun Janice Ahn, Wenpeng Yin","submitted_at":"2025-04-02T01:19:37Z","abstract_excerpt":"While the inconsistency of LLMs is not a novel topic, prior research has predominantly addressed two types of generative inconsistencies: i) Randomness Inconsistency: running the same LLM multiple trials, yielding varying responses; ii) Paraphrase Inconsistency: paraphrased prompts result in different responses from the same LLM. Randomness Inconsistency arises from the inherent randomness due to stochastic sampling in generative models, while Paraphrase Inconsistency is a consequence of the language modeling objectives, where paraphrased prompts alter the distribution of vocabulary logits. Th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.01282","kind":"arxiv","version":2},"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/2504.01282/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":"2504.01282","created_at":"2026-07-05T11:45:33.389779+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.01282v2","created_at":"2026-07-05T11:45:33.389779+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.01282","created_at":"2026-07-05T11:45:33.389779+00:00"},{"alias_kind":"pith_short_12","alias_value":"QP422M6HM6FF","created_at":"2026-07-05T11:45:33.389779+00:00"},{"alias_kind":"pith_short_16","alias_value":"QP422M6HM6FFCPS7","created_at":"2026-07-05T11:45:33.389779+00:00"},{"alias_kind":"pith_short_8","alias_value":"QP422M6H","created_at":"2026-07-05T11:45:33.389779+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2505.15134","citing_title":"The Unreasonable Effectiveness of Entropy Minimization in LLM Reasoning","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04852","citing_title":"Strengthening Human-Centric Chain-of-Thought Reasoning Integrity in LLMs via a Structured Prompt Framework","ref_index":30,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX","json":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX.json","graph_json":"https://pith.science/api/pith-number/QP422M6HM6FFCPS7JQPVUNLMWX/graph.json","events_json":"https://pith.science/api/pith-number/QP422M6HM6FFCPS7JQPVUNLMWX/events.json","paper":"https://pith.science/paper/QP422M6H"},"agent_actions":{"view_html":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX","download_json":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX.json","view_paper":"https://pith.science/paper/QP422M6H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.01282&json=true","fetch_graph":"https://pith.science/api/pith-number/QP422M6HM6FFCPS7JQPVUNLMWX/graph.json","fetch_events":"https://pith.science/api/pith-number/QP422M6HM6FFCPS7JQPVUNLMWX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX/action/storage_attestation","attest_author":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX/action/author_attestation","sign_citation":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX/action/citation_signature","submit_replication":"https://pith.science/pith/QP422M6HM6FFCPS7JQPVUNLMWX/action/replication_record"}},"created_at":"2026-07-05T11:45:33.389779+00:00","updated_at":"2026-07-05T11:45:33.389779+00:00"}