{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:ZN7PMRYFABJCKJWUGLHYASIXH2","short_pith_number":"pith:ZN7PMRYF","canonical_record":{"source":{"id":"2310.08008","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-10-12T03:20:43Z","cross_cats_sorted":[],"title_canon_sha256":"fcd9727c49dd1403e359b53a61bbc0325a1a3ed14622037cf176aa5b9b00e27c","abstract_canon_sha256":"54c50eb507b1bc522a0599222ecfefbbf24175e74ecd8e5adfb97b4d7e3cc758"},"schema_version":"1.0"},"canonical_sha256":"cb7ef6470500522526d432cf8049173e8e2f2157ee78d9e4e49fa3c7cd539e05","source":{"kind":"arxiv","id":"2310.08008","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.08008","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"arxiv_version","alias_value":"2310.08008v4","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.08008","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"pith_short_12","alias_value":"ZN7PMRYFABJC","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"pith_short_16","alias_value":"ZN7PMRYFABJCKJWU","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"pith_short_8","alias_value":"ZN7PMRYF","created_at":"2026-07-05T07:22:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:ZN7PMRYFABJCKJWUGLHYASIXH2","target":"record","payload":{"canonical_record":{"source":{"id":"2310.08008","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-10-12T03:20:43Z","cross_cats_sorted":[],"title_canon_sha256":"fcd9727c49dd1403e359b53a61bbc0325a1a3ed14622037cf176aa5b9b00e27c","abstract_canon_sha256":"54c50eb507b1bc522a0599222ecfefbbf24175e74ecd8e5adfb97b4d7e3cc758"},"schema_version":"1.0"},"canonical_sha256":"cb7ef6470500522526d432cf8049173e8e2f2157ee78d9e4e49fa3c7cd539e05","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:22:24.940352Z","signature_b64":"Ph9sfyQxUHM2XVLBaDZOWMNY2mF1CluH+oNKXzTrFGutzS6X2CAb1KxScaPAkMND9sADI+yVqzLxMbpm8va5DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cb7ef6470500522526d432cf8049173e8e2f2157ee78d9e4e49fa3c7cd539e05","last_reissued_at":"2026-07-05T07:22:24.939870Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:22:24.939870Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.08008","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:22:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ovynyuCb1rWKKzWUUAgiJ1iW6eOGlcKoasaSL9O98gW5iGZ/67AfzHXB1jM+7kpl+H52xPwyWNIxEwELqtS7Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T10:05:25.068367Z"},"content_sha256":"e9bf2009eb6c6b678de667d40ac1a8d31ab066f1804c787b6ad17fe1f5323979","schema_version":"1.0","event_id":"sha256:e9bf2009eb6c6b678de667d40ac1a8d31ab066f1804c787b6ad17fe1f5323979"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:ZN7PMRYFABJCKJWUGLHYASIXH2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Effects of Human Adversarial and Affable Samples on BERT Generalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Aparna Elangovan, Jiayuan He, Karin Verspoor, Yuan Li","submitted_at":"2023-10-12T03:20:43Z","abstract_excerpt":"BERT-based models have had strong performance on leaderboards, yet have been demonstrably worse in real-world settings requiring generalization. Limited quantities of training data is considered a key impediment to achieving generalizability in machine learning. In this paper, we examine the impact of training data quality, not quantity, on a model's generalizability. We consider two characteristics of training data: the portion of human-adversarial (h-adversarial), i.e., sample pairs with seemingly minor differences but different ground-truth labels, and human-affable (h-affable) training sam"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.08008","kind":"arxiv","version":4},"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/2310.08008/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":"2026-07-05T07:22:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gFkNso+r0nazF//hPaiGJkAW3z6lCIpp+wxiZUgc8fes9m+nXRL6tXAXr4ADbuW0QywudBr3R+Ev/gIKXfxWBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T10:05:25.068733Z"},"content_sha256":"ac1b1421c8049a789cb5ae00084294ceca4dde96f0c062ddf5f7fd93f5fa5fbc","schema_version":"1.0","event_id":"sha256:ac1b1421c8049a789cb5ae00084294ceca4dde96f0c062ddf5f7fd93f5fa5fbc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZN7PMRYFABJCKJWUGLHYASIXH2/bundle.json","state_url":"https://pith.science/pith/ZN7PMRYFABJCKJWUGLHYASIXH2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZN7PMRYFABJCKJWUGLHYASIXH2/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":"2026-07-06T10:05:25Z","links":{"resolver":"https://pith.science/pith/ZN7PMRYFABJCKJWUGLHYASIXH2","bundle":"https://pith.science/pith/ZN7PMRYFABJCKJWUGLHYASIXH2/bundle.json","state":"https://pith.science/pith/ZN7PMRYFABJCKJWUGLHYASIXH2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZN7PMRYFABJCKJWUGLHYASIXH2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:ZN7PMRYFABJCKJWUGLHYASIXH2","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":"54c50eb507b1bc522a0599222ecfefbbf24175e74ecd8e5adfb97b4d7e3cc758","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-10-12T03:20:43Z","title_canon_sha256":"fcd9727c49dd1403e359b53a61bbc0325a1a3ed14622037cf176aa5b9b00e27c"},"schema_version":"1.0","source":{"id":"2310.08008","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.08008","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"arxiv_version","alias_value":"2310.08008v4","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.08008","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"pith_short_12","alias_value":"ZN7PMRYFABJC","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"pith_short_16","alias_value":"ZN7PMRYFABJCKJWU","created_at":"2026-07-05T07:22:24Z"},{"alias_kind":"pith_short_8","alias_value":"ZN7PMRYF","created_at":"2026-07-05T07:22:24Z"}],"graph_snapshots":[{"event_id":"sha256:ac1b1421c8049a789cb5ae00084294ceca4dde96f0c062ddf5f7fd93f5fa5fbc","target":"graph","created_at":"2026-07-05T07:22:24Z","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/2310.08008/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"BERT-based models have had strong performance on leaderboards, yet have been demonstrably worse in real-world settings requiring generalization. Limited quantities of training data is considered a key impediment to achieving generalizability in machine learning. In this paper, we examine the impact of training data quality, not quantity, on a model's generalizability. We consider two characteristics of training data: the portion of human-adversarial (h-adversarial), i.e., sample pairs with seemingly minor differences but different ground-truth labels, and human-affable (h-affable) training sam","authors_text":"Aparna Elangovan, Jiayuan He, Karin Verspoor, Yuan Li","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-10-12T03:20:43Z","title":"Effects of Human Adversarial and Affable Samples on BERT Generalization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.08008","kind":"arxiv","version":4},"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:e9bf2009eb6c6b678de667d40ac1a8d31ab066f1804c787b6ad17fe1f5323979","target":"record","created_at":"2026-07-05T07:22:24Z","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":"54c50eb507b1bc522a0599222ecfefbbf24175e74ecd8e5adfb97b4d7e3cc758","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-10-12T03:20:43Z","title_canon_sha256":"fcd9727c49dd1403e359b53a61bbc0325a1a3ed14622037cf176aa5b9b00e27c"},"schema_version":"1.0","source":{"id":"2310.08008","kind":"arxiv","version":4}},"canonical_sha256":"cb7ef6470500522526d432cf8049173e8e2f2157ee78d9e4e49fa3c7cd539e05","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cb7ef6470500522526d432cf8049173e8e2f2157ee78d9e4e49fa3c7cd539e05","first_computed_at":"2026-07-05T07:22:24.939870Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:22:24.939870Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Ph9sfyQxUHM2XVLBaDZOWMNY2mF1CluH+oNKXzTrFGutzS6X2CAb1KxScaPAkMND9sADI+yVqzLxMbpm8va5DA==","signature_status":"signed_v1","signed_at":"2026-07-05T07:22:24.940352Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.08008","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e9bf2009eb6c6b678de667d40ac1a8d31ab066f1804c787b6ad17fe1f5323979","sha256:ac1b1421c8049a789cb5ae00084294ceca4dde96f0c062ddf5f7fd93f5fa5fbc"],"state_sha256":"314510e0a9410140fa5f14458e4f504185140e50d500997f738d1f6b02214230"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fgZA0kRKGXCoo4E0gVywpDeusPtXpRlKAcbBaVcAhrxwLuSXxvTivw67JektB+XDRssp+7qrR1JkjAGImQKzBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T10:05:25.070676Z","bundle_sha256":"29285fbb81fbbe4d16b6efe15c7e6257675534798913cb0c517fd5694e6402ac"}}