{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:6YR54WLZBGXTEGZ3IE56FE6Z46","short_pith_number":"pith:6YR54WLZ","canonical_record":{"source":{"id":"1907.04380","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-07-09T19:40:47Z","cross_cats_sorted":[],"title_canon_sha256":"13d518f385bb9e16c7193380de7d567cf9bc6553be08d180ddd5ea4599ceaf22","abstract_canon_sha256":"d33826c6aaeb77e5ad2b1ec9d8729b30b1ed956cfe87a383c19f9b9ca904979c"},"schema_version":"1.0"},"canonical_sha256":"f623de597909af321b3b413be293d9e7a3bc36ded91e12305c4d4df04d175f07","source":{"kind":"arxiv","id":"1907.04380","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1907.04380","created_at":"2026-05-17T23:41:01Z"},{"alias_kind":"arxiv_version","alias_value":"1907.04380v1","created_at":"2026-05-17T23:41:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1907.04380","created_at":"2026-05-17T23:41:01Z"},{"alias_kind":"pith_short_12","alias_value":"6YR54WLZBGXT","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_16","alias_value":"6YR54WLZBGXTEGZ3","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_8","alias_value":"6YR54WLZ","created_at":"2026-05-18T12:33:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:6YR54WLZBGXTEGZ3IE56FE6Z46","target":"record","payload":{"canonical_record":{"source":{"id":"1907.04380","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-07-09T19:40:47Z","cross_cats_sorted":[],"title_canon_sha256":"13d518f385bb9e16c7193380de7d567cf9bc6553be08d180ddd5ea4599ceaf22","abstract_canon_sha256":"d33826c6aaeb77e5ad2b1ec9d8729b30b1ed956cfe87a383c19f9b9ca904979c"},"schema_version":"1.0"},"canonical_sha256":"f623de597909af321b3b413be293d9e7a3bc36ded91e12305c4d4df04d175f07","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:41:01.934844Z","signature_b64":"Fidg7Nngoof8gsdqiSimtpCBKodiz7QwnrS63aBY0PmtNsogL5ZC0Xq6LT18l4xF23mK10hZiICcPeh/M+10Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f623de597909af321b3b413be293d9e7a3bc36ded91e12305c4d4df04d175f07","last_reissued_at":"2026-05-17T23:41:01.934047Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:41:01.934047Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1907.04380","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":"2026-05-17T23:41:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dDuGIHV8kIpBGtuDqVZ8ZNcsB8HP3J0MQ7jQ8aG+g+7qquPaL3QSpdz1kLGDh6w2ic3AzmMQB834OsMRv3YtBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T22:36:31.414875Z"},"content_sha256":"02769b6de2bdd53c35d5f2ff6438f95659e6f3cdee9568f7b3c15fac4265f8ed","schema_version":"1.0","event_id":"sha256:02769b6de2bdd53c35d5f2ff6438f95659e6f3cdee9568f7b3c15fac4265f8ed"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:6YR54WLZBGXTEGZ3IE56FE6Z46","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Don't Take the Premise for Granted: Mitigating Artifacts in Natural Language Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Adam Poliak, Alexander M. Rush, Benjamin Van Durme, Stuart M. Shieber, Yonatan Belinkov","submitted_at":"2019-07-09T19:40:47Z","abstract_excerpt":"Natural Language Inference (NLI) datasets often contain hypothesis-only biases---artifacts that allow models to achieve non-trivial performance without learning whether a premise entails a hypothesis. We propose two probabilistic methods to build models that are more robust to such biases and better transfer across datasets. In contrast to standard approaches to NLI, our methods predict the probability of a premise given a hypothesis and NLI label, discouraging models from ignoring the premise. We evaluate our methods on synthetic and existing NLI datasets by training on datasets containing bi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1907.04380","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":""},"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":"2026-05-17T23:41:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"d49F46frrg6+esjHS63dn8dzaOf7nZMhYinVU2496uEPEq0qlfP9ZJGPYXCYZt37keTVJSaT8HaCK0KIVm5AAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T22:36:31.415604Z"},"content_sha256":"7eeb168df23ae28e3a1465975897c0dad9f2d4ea781883f314d7cbe9022e40bb","schema_version":"1.0","event_id":"sha256:7eeb168df23ae28e3a1465975897c0dad9f2d4ea781883f314d7cbe9022e40bb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6YR54WLZBGXTEGZ3IE56FE6Z46/bundle.json","state_url":"https://pith.science/pith/6YR54WLZBGXTEGZ3IE56FE6Z46/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6YR54WLZBGXTEGZ3IE56FE6Z46/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":"2026-05-26T22:36:31Z","links":{"resolver":"https://pith.science/pith/6YR54WLZBGXTEGZ3IE56FE6Z46","bundle":"https://pith.science/pith/6YR54WLZBGXTEGZ3IE56FE6Z46/bundle.json","state":"https://pith.science/pith/6YR54WLZBGXTEGZ3IE56FE6Z46/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6YR54WLZBGXTEGZ3IE56FE6Z46/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:6YR54WLZBGXTEGZ3IE56FE6Z46","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":"d33826c6aaeb77e5ad2b1ec9d8729b30b1ed956cfe87a383c19f9b9ca904979c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-07-09T19:40:47Z","title_canon_sha256":"13d518f385bb9e16c7193380de7d567cf9bc6553be08d180ddd5ea4599ceaf22"},"schema_version":"1.0","source":{"id":"1907.04380","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1907.04380","created_at":"2026-05-17T23:41:01Z"},{"alias_kind":"arxiv_version","alias_value":"1907.04380v1","created_at":"2026-05-17T23:41:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1907.04380","created_at":"2026-05-17T23:41:01Z"},{"alias_kind":"pith_short_12","alias_value":"6YR54WLZBGXT","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_16","alias_value":"6YR54WLZBGXTEGZ3","created_at":"2026-05-18T12:33:10Z"},{"alias_kind":"pith_short_8","alias_value":"6YR54WLZ","created_at":"2026-05-18T12:33:10Z"}],"graph_snapshots":[{"event_id":"sha256:7eeb168df23ae28e3a1465975897c0dad9f2d4ea781883f314d7cbe9022e40bb","target":"graph","created_at":"2026-05-17T23:41:01Z","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"},"paper":{"abstract_excerpt":"Natural Language Inference (NLI) datasets often contain hypothesis-only biases---artifacts that allow models to achieve non-trivial performance without learning whether a premise entails a hypothesis. We propose two probabilistic methods to build models that are more robust to such biases and better transfer across datasets. In contrast to standard approaches to NLI, our methods predict the probability of a premise given a hypothesis and NLI label, discouraging models from ignoring the premise. We evaluate our methods on synthetic and existing NLI datasets by training on datasets containing bi","authors_text":"Adam Poliak, Alexander M. Rush, Benjamin Van Durme, Stuart M. Shieber, Yonatan Belinkov","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-07-09T19:40:47Z","title":"Don't Take the Premise for Granted: Mitigating Artifacts in Natural Language Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1907.04380","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":"sha256:02769b6de2bdd53c35d5f2ff6438f95659e6f3cdee9568f7b3c15fac4265f8ed","target":"record","created_at":"2026-05-17T23:41:01Z","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":"d33826c6aaeb77e5ad2b1ec9d8729b30b1ed956cfe87a383c19f9b9ca904979c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-07-09T19:40:47Z","title_canon_sha256":"13d518f385bb9e16c7193380de7d567cf9bc6553be08d180ddd5ea4599ceaf22"},"schema_version":"1.0","source":{"id":"1907.04380","kind":"arxiv","version":1}},"canonical_sha256":"f623de597909af321b3b413be293d9e7a3bc36ded91e12305c4d4df04d175f07","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f623de597909af321b3b413be293d9e7a3bc36ded91e12305c4d4df04d175f07","first_computed_at":"2026-05-17T23:41:01.934047Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:41:01.934047Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Fidg7Nngoof8gsdqiSimtpCBKodiz7QwnrS63aBY0PmtNsogL5ZC0Xq6LT18l4xF23mK10hZiICcPeh/M+10Aw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:41:01.934844Z","signed_message":"canonical_sha256_bytes"},"source_id":"1907.04380","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:02769b6de2bdd53c35d5f2ff6438f95659e6f3cdee9568f7b3c15fac4265f8ed","sha256:7eeb168df23ae28e3a1465975897c0dad9f2d4ea781883f314d7cbe9022e40bb"],"state_sha256":"0e2116dd84a5e04674f8d4b75c1831462d2adf488c9ad46d2f8517a0f722522b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1Q8h23X+w4V3upHmlw3pTR5FTVRipOOjRHHsdUMtcI0bp139w3AFWmilbYlerfS/v/74T9//qPeYQ4+XVLhhAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T22:36:31.419180Z","bundle_sha256":"27cc9b29fe27beb423fef2b7c78c4a90c6ff62f76050a3eb79334c88ca5dcbf9"}}