{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:OWQT7A5IZBNJM2P7G47S35VT2E","short_pith_number":"pith:OWQT7A5I","canonical_record":{"source":{"id":"2309.07822","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-14T16:16:40Z","cross_cats_sorted":[],"title_canon_sha256":"c2327eceffe9fb3ee781bc1e1e9425e12b91709d877ee038ea522f2c2b14c739","abstract_canon_sha256":"61361fc9d1e2819b2c639e4ef989c02c4bf7baeb6091ecc9a9de35cf1690d456"},"schema_version":"1.0"},"canonical_sha256":"75a13f83a8c85a9669ff373f2df6b3d130a41828a5b5d0918f5c6062f4f1b24e","source":{"kind":"arxiv","id":"2309.07822","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.07822","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"arxiv_version","alias_value":"2309.07822v3","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.07822","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"pith_short_12","alias_value":"OWQT7A5IZBNJ","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"pith_short_16","alias_value":"OWQT7A5IZBNJM2P7","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"pith_short_8","alias_value":"OWQT7A5I","created_at":"2026-07-05T07:44:20Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:OWQT7A5IZBNJM2P7G47S35VT2E","target":"record","payload":{"canonical_record":{"source":{"id":"2309.07822","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-14T16:16:40Z","cross_cats_sorted":[],"title_canon_sha256":"c2327eceffe9fb3ee781bc1e1e9425e12b91709d877ee038ea522f2c2b14c739","abstract_canon_sha256":"61361fc9d1e2819b2c639e4ef989c02c4bf7baeb6091ecc9a9de35cf1690d456"},"schema_version":"1.0"},"canonical_sha256":"75a13f83a8c85a9669ff373f2df6b3d130a41828a5b5d0918f5c6062f4f1b24e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:44:20.337977Z","signature_b64":"W6y86yrBS68iTS9X/DrG92QX8Gf5pg2uTayobkk/+OMjaIumi1H4HW2fsaw1SFYr0F+Y8EUEnWTEMeiB3EsTAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"75a13f83a8c85a9669ff373f2df6b3d130a41828a5b5d0918f5c6062f4f1b24e","last_reissued_at":"2026-07-05T07:44:20.337469Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:44:20.337469Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2309.07822","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:44:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zGQFxB1JT+WxoAo/goo1IumwBnchFZf8hnr39b+KYUqRG7l7u3QDodEq+F8oNi5Iy2p3C9/aGu1zlb3EoN7ZDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T19:30:18.852041Z"},"content_sha256":"4c84ae5680b4ac0bfc4f60f96b73670ee356cd404520a69ca8ab7d572ba6c4e0","schema_version":"1.0","event_id":"sha256:4c84ae5680b4ac0bfc4f60f96b73670ee356cd404520a69ca8ab7d572ba6c4e0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:OWQT7A5IZBNJM2P7G47S35VT2E","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CATfOOD: Counterfactual Augmented Training for Improving Out-of-Domain Performance and Calibration","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Iryna Gurevych, Martin Tutek, Rachneet Sachdeva","submitted_at":"2023-09-14T16:16:40Z","abstract_excerpt":"In recent years, large language models (LLMs) have shown remarkable capabilities at scale, particularly at generating text conditioned on a prompt. In our work, we investigate the use of LLMs to augment training data of small language models~(SLMs) with automatically generated counterfactual~(CF) instances -- i.e. minimally altered inputs -- in order to improve out-of-domain~(OOD) performance of SLMs in the extractive question answering~(QA) setup. We show that, across various LLM generators, such data augmentation consistently enhances OOD performance and improves model calibration for both c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.07822","kind":"arxiv","version":3},"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/2309.07822/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":"2026-07-05T07:44:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hktnX06LuTrv9vr/1AMMC2DMLuStNcHb+OLPkM8z5du03fPXLAH/EIzXywQTWbh1Rw3oToUCMD0DmOlXWQCWDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T19:30:18.852691Z"},"content_sha256":"aae40ffce28eba7ab85a2df37970a60e6b024a021ae793e84f46ec49f1eaed7a","schema_version":"1.0","event_id":"sha256:aae40ffce28eba7ab85a2df37970a60e6b024a021ae793e84f46ec49f1eaed7a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OWQT7A5IZBNJM2P7G47S35VT2E/bundle.json","state_url":"https://pith.science/pith/OWQT7A5IZBNJM2P7G47S35VT2E/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OWQT7A5IZBNJM2P7G47S35VT2E/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":"2026-08-17T19:30:18Z","links":{"resolver":"https://pith.science/pith/OWQT7A5IZBNJM2P7G47S35VT2E","bundle":"https://pith.science/pith/OWQT7A5IZBNJM2P7G47S35VT2E/bundle.json","state":"https://pith.science/pith/OWQT7A5IZBNJM2P7G47S35VT2E/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OWQT7A5IZBNJM2P7G47S35VT2E/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:OWQT7A5IZBNJM2P7G47S35VT2E","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":"61361fc9d1e2819b2c639e4ef989c02c4bf7baeb6091ecc9a9de35cf1690d456","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-14T16:16:40Z","title_canon_sha256":"c2327eceffe9fb3ee781bc1e1e9425e12b91709d877ee038ea522f2c2b14c739"},"schema_version":"1.0","source":{"id":"2309.07822","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.07822","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"arxiv_version","alias_value":"2309.07822v3","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.07822","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"pith_short_12","alias_value":"OWQT7A5IZBNJ","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"pith_short_16","alias_value":"OWQT7A5IZBNJM2P7","created_at":"2026-07-05T07:44:20Z"},{"alias_kind":"pith_short_8","alias_value":"OWQT7A5I","created_at":"2026-07-05T07:44:20Z"}],"graph_snapshots":[{"event_id":"sha256:aae40ffce28eba7ab85a2df37970a60e6b024a021ae793e84f46ec49f1eaed7a","target":"graph","created_at":"2026-07-05T07:44:20Z","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/2309.07822/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In recent years, large language models (LLMs) have shown remarkable capabilities at scale, particularly at generating text conditioned on a prompt. In our work, we investigate the use of LLMs to augment training data of small language models~(SLMs) with automatically generated counterfactual~(CF) instances -- i.e. minimally altered inputs -- in order to improve out-of-domain~(OOD) performance of SLMs in the extractive question answering~(QA) setup. We show that, across various LLM generators, such data augmentation consistently enhances OOD performance and improves model calibration for both c","authors_text":"Iryna Gurevych, Martin Tutek, Rachneet Sachdeva","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-14T16:16:40Z","title":"CATfOOD: Counterfactual Augmented Training for Improving Out-of-Domain Performance and Calibration"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.07822","kind":"arxiv","version":3},"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":"sha256:4c84ae5680b4ac0bfc4f60f96b73670ee356cd404520a69ca8ab7d572ba6c4e0","target":"record","created_at":"2026-07-05T07:44:20Z","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":"61361fc9d1e2819b2c639e4ef989c02c4bf7baeb6091ecc9a9de35cf1690d456","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-14T16:16:40Z","title_canon_sha256":"c2327eceffe9fb3ee781bc1e1e9425e12b91709d877ee038ea522f2c2b14c739"},"schema_version":"1.0","source":{"id":"2309.07822","kind":"arxiv","version":3}},"canonical_sha256":"75a13f83a8c85a9669ff373f2df6b3d130a41828a5b5d0918f5c6062f4f1b24e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"75a13f83a8c85a9669ff373f2df6b3d130a41828a5b5d0918f5c6062f4f1b24e","first_computed_at":"2026-07-05T07:44:20.337469Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:44:20.337469Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"W6y86yrBS68iTS9X/DrG92QX8Gf5pg2uTayobkk/+OMjaIumi1H4HW2fsaw1SFYr0F+Y8EUEnWTEMeiB3EsTAg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:44:20.337977Z","signed_message":"canonical_sha256_bytes"},"source_id":"2309.07822","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4c84ae5680b4ac0bfc4f60f96b73670ee356cd404520a69ca8ab7d572ba6c4e0","sha256:aae40ffce28eba7ab85a2df37970a60e6b024a021ae793e84f46ec49f1eaed7a"],"state_sha256":"b044f7289d320fc2348ba5e754927d8d6dd933065e09317c77c4fa1c03d12157"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yypxiCG0jTC+o4/TN++cIISX5GmUXD7ku/T16WWtv7FllUSF5UPp6Yptiu2tYWhee1waEN8yP74CL9UXbCMXCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T19:30:18.860070Z","bundle_sha256":"508c7682f71a4984f4c99297a390f145419b3e198cb77d8b936dea867c2107e9"}}