{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:LLJAPIKYQ6KXEXH6SZORPI45DG","short_pith_number":"pith:LLJAPIKY","canonical_record":{"source":{"id":"2405.00722","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-26T11:57:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"67d6e3407d707ac723c4093d341b13fc16026d515544d205728fe330fbd075b2","abstract_canon_sha256":"42bcbd5f50bc9ab4cc5e9881f132687c24ce9a7183f0f6c4c8b67d78045bb3a0"},"schema_version":"1.0"},"canonical_sha256":"5ad207a1588795725cfe965d17a39d1986009f5b93bb140f19ba0ecfc27dd878","source":{"kind":"arxiv","id":"2405.00722","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.00722","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"arxiv_version","alias_value":"2405.00722v2","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.00722","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_12","alias_value":"LLJAPIKYQ6KX","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_16","alias_value":"LLJAPIKYQ6KXEXH6","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_8","alias_value":"LLJAPIKY","created_at":"2026-07-05T09:34:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:LLJAPIKYQ6KXEXH6SZORPI45DG","target":"record","payload":{"canonical_record":{"source":{"id":"2405.00722","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-26T11:57:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"67d6e3407d707ac723c4093d341b13fc16026d515544d205728fe330fbd075b2","abstract_canon_sha256":"42bcbd5f50bc9ab4cc5e9881f132687c24ce9a7183f0f6c4c8b67d78045bb3a0"},"schema_version":"1.0"},"canonical_sha256":"5ad207a1588795725cfe965d17a39d1986009f5b93bb140f19ba0ecfc27dd878","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:34:13.443332Z","signature_b64":"oVdj1sIVlDfTzNWCTUY66soXahrhoTEG9KqU3gs/ipS4QIfgRhDnvXI9vUa+SOv3LFyWEHVzRMSA4gptpUN5Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5ad207a1588795725cfe965d17a39d1986009f5b93bb140f19ba0ecfc27dd878","last_reissued_at":"2026-07-05T09:34:13.442895Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:34:13.442895Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.00722","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:34:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2hxp/BD6qKye1wq1aSyKeUcsjoxIQ6yjgd5xbe7gVCioe1ahRBPu/UBt11bNyegScbcIaNyV0CZBRXtre0Z9Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T15:45:44.229752Z"},"content_sha256":"d483b8254d46ac18a3fc5ae050cd9a22660cb754a566a7fc15e5c45e97a2e12f","schema_version":"1.0","event_id":"sha256:d483b8254d46ac18a3fc5ae050cd9a22660cb754a566a7fc15e5c45e97a2e12f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:LLJAPIKYQ6KXEXH6SZORPI45DG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LLMs for Generating and Evaluating Counterfactuals: A Comprehensive Study","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Christin Seifert, J\\\"org Schl\\\"otterer, Paul Youssef, Van Bach Nguyen","submitted_at":"2024-04-26T11:57:21Z","abstract_excerpt":"As NLP models become more complex, understanding their decisions becomes more crucial. Counterfactuals (CFs), where minimal changes to inputs flip a model's prediction, offer a way to explain these models. While Large Language Models (LLMs) have shown remarkable performance in NLP tasks, their efficacy in generating high-quality CFs remains uncertain. This work fills this gap by investigating how well LLMs generate CFs for two NLU tasks. We conduct a comprehensive comparison of several common LLMs, and evaluate their CFs, assessing both intrinsic metrics, and the impact of these CFs on data au"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.00722","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/2405.00722/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":"2026-07-05T09:34:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mZIPGfhMYkv5zsWVb3Sompkq4sozth/aVIzXd88aABxOn2wwUM2P0sKw2goV3FBlhzeWRZI2LTDo8jLVz8ZCCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T15:45:44.230724Z"},"content_sha256":"21d644c10ce3ff521a390fa99bf7501d2d5b5a4f90e9a00393adfe9da613b75e","schema_version":"1.0","event_id":"sha256:21d644c10ce3ff521a390fa99bf7501d2d5b5a4f90e9a00393adfe9da613b75e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LLJAPIKYQ6KXEXH6SZORPI45DG/bundle.json","state_url":"https://pith.science/pith/LLJAPIKYQ6KXEXH6SZORPI45DG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LLJAPIKYQ6KXEXH6SZORPI45DG/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":"2026-08-13T15:45:44Z","links":{"resolver":"https://pith.science/pith/LLJAPIKYQ6KXEXH6SZORPI45DG","bundle":"https://pith.science/pith/LLJAPIKYQ6KXEXH6SZORPI45DG/bundle.json","state":"https://pith.science/pith/LLJAPIKYQ6KXEXH6SZORPI45DG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LLJAPIKYQ6KXEXH6SZORPI45DG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:LLJAPIKYQ6KXEXH6SZORPI45DG","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":"42bcbd5f50bc9ab4cc5e9881f132687c24ce9a7183f0f6c4c8b67d78045bb3a0","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-26T11:57:21Z","title_canon_sha256":"67d6e3407d707ac723c4093d341b13fc16026d515544d205728fe330fbd075b2"},"schema_version":"1.0","source":{"id":"2405.00722","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.00722","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"arxiv_version","alias_value":"2405.00722v2","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.00722","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_12","alias_value":"LLJAPIKYQ6KX","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_16","alias_value":"LLJAPIKYQ6KXEXH6","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_8","alias_value":"LLJAPIKY","created_at":"2026-07-05T09:34:13Z"}],"graph_snapshots":[{"event_id":"sha256:21d644c10ce3ff521a390fa99bf7501d2d5b5a4f90e9a00393adfe9da613b75e","target":"graph","created_at":"2026-07-05T09:34:13Z","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/2405.00722/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As NLP models become more complex, understanding their decisions becomes more crucial. Counterfactuals (CFs), where minimal changes to inputs flip a model's prediction, offer a way to explain these models. While Large Language Models (LLMs) have shown remarkable performance in NLP tasks, their efficacy in generating high-quality CFs remains uncertain. This work fills this gap by investigating how well LLMs generate CFs for two NLU tasks. We conduct a comprehensive comparison of several common LLMs, and evaluate their CFs, assessing both intrinsic metrics, and the impact of these CFs on data au","authors_text":"Christin Seifert, J\\\"org Schl\\\"otterer, Paul Youssef, Van Bach Nguyen","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-26T11:57:21Z","title":"LLMs for Generating and Evaluating Counterfactuals: A Comprehensive Study"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.00722","kind":"arxiv","version":2},"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:d483b8254d46ac18a3fc5ae050cd9a22660cb754a566a7fc15e5c45e97a2e12f","target":"record","created_at":"2026-07-05T09:34:13Z","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":"42bcbd5f50bc9ab4cc5e9881f132687c24ce9a7183f0f6c4c8b67d78045bb3a0","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-26T11:57:21Z","title_canon_sha256":"67d6e3407d707ac723c4093d341b13fc16026d515544d205728fe330fbd075b2"},"schema_version":"1.0","source":{"id":"2405.00722","kind":"arxiv","version":2}},"canonical_sha256":"5ad207a1588795725cfe965d17a39d1986009f5b93bb140f19ba0ecfc27dd878","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5ad207a1588795725cfe965d17a39d1986009f5b93bb140f19ba0ecfc27dd878","first_computed_at":"2026-07-05T09:34:13.442895Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:34:13.442895Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oVdj1sIVlDfTzNWCTUY66soXahrhoTEG9KqU3gs/ipS4QIfgRhDnvXI9vUa+SOv3LFyWEHVzRMSA4gptpUN5Dw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:34:13.443332Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.00722","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d483b8254d46ac18a3fc5ae050cd9a22660cb754a566a7fc15e5c45e97a2e12f","sha256:21d644c10ce3ff521a390fa99bf7501d2d5b5a4f90e9a00393adfe9da613b75e"],"state_sha256":"442aa935a907ca9252d61aacfc642cdb74bafaa762afa364890554fbaf07a9ad"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4wgcZZt7GdrmA35tXA9ZezNzElj9jcmsiJbd9Q/TmR1MhsGyE13umySaBiZFpP7fCLSeA0K4A0wG6eqw+AjADw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-13T15:45:44.235595Z","bundle_sha256":"9947f55737bf97ce062d33fb581cad269a70a09d6a1fb7583880bfceb0383d9a"}}