{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:Z4AHT6JA73PCS46HUVTGNQWGLY","short_pith_number":"pith:Z4AHT6JA","canonical_record":{"source":{"id":"2311.17107","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-11-28T10:26:57Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CY","cs.IR"],"title_canon_sha256":"5b7f6b3bc65d8d5dfe495ea1d37b297d5d244d212a5fa501ae95447c02cd026e","abstract_canon_sha256":"64cca5d188c6667420fb828b4b412e8bd7ee3158a2977708b2f50750a728494c"},"schema_version":"1.0"},"canonical_sha256":"cf0079f920fede2973c7a56666c2c65e2203c7b04ce1b0758e51c345e0b69899","source":{"kind":"arxiv","id":"2311.17107","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.17107","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"arxiv_version","alias_value":"2311.17107v1","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.17107","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"pith_short_12","alias_value":"Z4AHT6JA73PC","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"pith_short_16","alias_value":"Z4AHT6JA73PCS46H","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"pith_short_8","alias_value":"Z4AHT6JA","created_at":"2026-07-05T07:17:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:Z4AHT6JA73PCS46HUVTGNQWGLY","target":"record","payload":{"canonical_record":{"source":{"id":"2311.17107","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-11-28T10:26:57Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CY","cs.IR"],"title_canon_sha256":"5b7f6b3bc65d8d5dfe495ea1d37b297d5d244d212a5fa501ae95447c02cd026e","abstract_canon_sha256":"64cca5d188c6667420fb828b4b412e8bd7ee3158a2977708b2f50750a728494c"},"schema_version":"1.0"},"canonical_sha256":"cf0079f920fede2973c7a56666c2c65e2203c7b04ce1b0758e51c345e0b69899","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:17:57.569345Z","signature_b64":"NQi/NADxlaYot2+nJejDEBZGxq2x1ElTeferHWmR3RrKni1RFDaAR6H3TumLIUvMDFSGANPtloM0SlRm2DzpAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cf0079f920fede2973c7a56666c2c65e2203c7b04ce1b0758e51c345e0b69899","last_reissued_at":"2026-07-05T07:17:57.568857Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:17:57.568857Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2311.17107","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:17:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"flvvek7idF6KIW3qtHyYeRirEV1/+jRzKHO4qvLrfpsKeJVyfrSOoMaBd5yddacxsy58CI/Ly8OLGOb0of9gBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-23T05:17:20.732640Z"},"content_sha256":"6b8963a208c2ef40b961d71b27f508409866329a6ee03a9bf1b9c73c265e60a2","schema_version":"1.0","event_id":"sha256:6b8963a208c2ef40b961d71b27f508409866329a6ee03a9bf1b9c73c265e60a2"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:Z4AHT6JA73PCS46HUVTGNQWGLY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ClimateX: Do LLMs Accurately Assess Human Expert Confidence in Climate Statements?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CY","cs.IR"],"primary_cat":"cs.LG","authors_text":"Eddie Dilworth, Kerrie Wu, Romain Lacombe","submitted_at":"2023-11-28T10:26:57Z","abstract_excerpt":"Evaluating the accuracy of outputs generated by Large Language Models (LLMs) is especially important in the climate science and policy domain. We introduce the Expert Confidence in Climate Statements (ClimateX) dataset, a novel, curated, expert-labeled dataset consisting of 8094 climate statements collected from the latest Intergovernmental Panel on Climate Change (IPCC) reports, labeled with their associated confidence levels. Using this dataset, we show that recent LLMs can classify human expert confidence in climate-related statements, especially in a few-shot learning setting, but with lim"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.17107","kind":"arxiv","version":1},"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/2311.17107/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:17:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cM07ZAwJ5pk5TSc+yH9Yky8v4/tesQq7cl2ECy1z91JIoN6xA939N3eSgcQSrpy1Yb0l5elwsHe1NtoEAW8gDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-23T05:17:20.733129Z"},"content_sha256":"2d6802a6b52d7e623c392fa1834a9c14e5efcdcf45ba49c3fb34ec663f593145","schema_version":"1.0","event_id":"sha256:2d6802a6b52d7e623c392fa1834a9c14e5efcdcf45ba49c3fb34ec663f593145"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Z4AHT6JA73PCS46HUVTGNQWGLY/bundle.json","state_url":"https://pith.science/pith/Z4AHT6JA73PCS46HUVTGNQWGLY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Z4AHT6JA73PCS46HUVTGNQWGLY/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-23T05:17:20Z","links":{"resolver":"https://pith.science/pith/Z4AHT6JA73PCS46HUVTGNQWGLY","bundle":"https://pith.science/pith/Z4AHT6JA73PCS46HUVTGNQWGLY/bundle.json","state":"https://pith.science/pith/Z4AHT6JA73PCS46HUVTGNQWGLY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Z4AHT6JA73PCS46HUVTGNQWGLY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:Z4AHT6JA73PCS46HUVTGNQWGLY","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":"64cca5d188c6667420fb828b4b412e8bd7ee3158a2977708b2f50750a728494c","cross_cats_sorted":["cs.AI","cs.CL","cs.CY","cs.IR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-11-28T10:26:57Z","title_canon_sha256":"5b7f6b3bc65d8d5dfe495ea1d37b297d5d244d212a5fa501ae95447c02cd026e"},"schema_version":"1.0","source":{"id":"2311.17107","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.17107","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"arxiv_version","alias_value":"2311.17107v1","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.17107","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"pith_short_12","alias_value":"Z4AHT6JA73PC","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"pith_short_16","alias_value":"Z4AHT6JA73PCS46H","created_at":"2026-07-05T07:17:57Z"},{"alias_kind":"pith_short_8","alias_value":"Z4AHT6JA","created_at":"2026-07-05T07:17:57Z"}],"graph_snapshots":[{"event_id":"sha256:2d6802a6b52d7e623c392fa1834a9c14e5efcdcf45ba49c3fb34ec663f593145","target":"graph","created_at":"2026-07-05T07:17:57Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2311.17107/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluating the accuracy of outputs generated by Large Language Models (LLMs) is especially important in the climate science and policy domain. We introduce the Expert Confidence in Climate Statements (ClimateX) dataset, a novel, curated, expert-labeled dataset consisting of 8094 climate statements collected from the latest Intergovernmental Panel on Climate Change (IPCC) reports, labeled with their associated confidence levels. Using this dataset, we show that recent LLMs can classify human expert confidence in climate-related statements, especially in a few-shot learning setting, but with lim","authors_text":"Eddie Dilworth, Kerrie Wu, Romain Lacombe","cross_cats":["cs.AI","cs.CL","cs.CY","cs.IR"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-11-28T10:26:57Z","title":"ClimateX: Do LLMs Accurately Assess Human Expert Confidence in Climate Statements?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.17107","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:6b8963a208c2ef40b961d71b27f508409866329a6ee03a9bf1b9c73c265e60a2","target":"record","created_at":"2026-07-05T07:17:57Z","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":"64cca5d188c6667420fb828b4b412e8bd7ee3158a2977708b2f50750a728494c","cross_cats_sorted":["cs.AI","cs.CL","cs.CY","cs.IR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-11-28T10:26:57Z","title_canon_sha256":"5b7f6b3bc65d8d5dfe495ea1d37b297d5d244d212a5fa501ae95447c02cd026e"},"schema_version":"1.0","source":{"id":"2311.17107","kind":"arxiv","version":1}},"canonical_sha256":"cf0079f920fede2973c7a56666c2c65e2203c7b04ce1b0758e51c345e0b69899","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cf0079f920fede2973c7a56666c2c65e2203c7b04ce1b0758e51c345e0b69899","first_computed_at":"2026-07-05T07:17:57.568857Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:17:57.568857Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"NQi/NADxlaYot2+nJejDEBZGxq2x1ElTeferHWmR3RrKni1RFDaAR6H3TumLIUvMDFSGANPtloM0SlRm2DzpAw==","signature_status":"signed_v1","signed_at":"2026-07-05T07:17:57.569345Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.17107","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6b8963a208c2ef40b961d71b27f508409866329a6ee03a9bf1b9c73c265e60a2","sha256:2d6802a6b52d7e623c392fa1834a9c14e5efcdcf45ba49c3fb34ec663f593145"],"state_sha256":"64e22829d9fc42bc310c25fcede8dbd84ef8f2df7301bb6704c25f95a254b236"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VBz94SbODoFLLUOKB8U7aSlzN99mKLSKujPyiUTCbYF/iLigDyHsQKYo9Fils1zByBqmjNx/aX++RMWTJaheBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-23T05:17:20.736898Z","bundle_sha256":"05b0c2382ab045f4e7da03d4e4ae57ca830f651672090088c40a07f5f251adc3"}}