{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:SGFSQFHKD7VMM6VNFJJW3KCCX4","short_pith_number":"pith:SGFSQFHK","canonical_record":{"source":{"id":"2208.07316","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2022-08-15T16:30:14Z","cross_cats_sorted":["cs.CR","cs.LG"],"title_canon_sha256":"a9a068e80a1948e286cd6fcd4d4b3b47aaa208004b8e99dff968f294033f42ca","abstract_canon_sha256":"6c79a1a480473c58492f288d5d4db3e1f07f2b88e6abb9d22524712713498a83"},"schema_version":"1.0"},"canonical_sha256":"918b2814ea1feac67aad2a536da842bf146c73e67c8ac4b4697549779c1eb37b","source":{"kind":"arxiv","id":"2208.07316","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2208.07316","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"arxiv_version","alias_value":"2208.07316v5","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.07316","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"pith_short_12","alias_value":"SGFSQFHKD7VM","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"pith_short_16","alias_value":"SGFSQFHKD7VMM6VN","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"pith_short_8","alias_value":"SGFSQFHK","created_at":"2026-07-05T07:27:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:SGFSQFHKD7VMM6VNFJJW3KCCX4","target":"record","payload":{"canonical_record":{"source":{"id":"2208.07316","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2022-08-15T16:30:14Z","cross_cats_sorted":["cs.CR","cs.LG"],"title_canon_sha256":"a9a068e80a1948e286cd6fcd4d4b3b47aaa208004b8e99dff968f294033f42ca","abstract_canon_sha256":"6c79a1a480473c58492f288d5d4db3e1f07f2b88e6abb9d22524712713498a83"},"schema_version":"1.0"},"canonical_sha256":"918b2814ea1feac67aad2a536da842bf146c73e67c8ac4b4697549779c1eb37b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:27:41.131250Z","signature_b64":"UPxOjYu18uKXc4SNS+W6yuSQeEfLJRiuJWc86MTjjZSj/cwM1w2FW7RRoH+hu6IVD0DKpOcGWpbxPMx7yLf9BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"918b2814ea1feac67aad2a536da842bf146c73e67c8ac4b4697549779c1eb37b","last_reissued_at":"2026-07-05T07:27:41.130770Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:27:41.130770Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2208.07316","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:27:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GRKwFL43xOt95t096nJda0CIF4XihwPZAjKDYIqj/zLJyeqviGDaG/AVViBeIXexoTVDWwORkvlCfpsqDlBqCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T22:36:59.868246Z"},"content_sha256":"bbabc8a2f7f65050b8beebcecff2cc167c86925ff095d99ac082453631e69339","schema_version":"1.0","event_id":"sha256:bbabc8a2f7f65050b8beebcecff2cc167c86925ff095d99ac082453631e69339"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:SGFSQFHKD7VMM6VNFJJW3KCCX4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MENLI: Robust Evaluation Metrics from Natural Language Inference","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Steffen Eger, Yanran Chen","submitted_at":"2022-08-15T16:30:14Z","abstract_excerpt":"Recently proposed BERT-based evaluation metrics for text generation perform well on standard benchmarks but are vulnerable to adversarial attacks, e.g., relating to information correctness. We argue that this stems (in part) from the fact that they are models of semantic similarity. In contrast, we develop evaluation metrics based on Natural Language Inference (NLI), which we deem a more appropriate modeling. We design a preference-based adversarial attack framework and show that our NLI based metrics are much more robust to the attacks than the recent BERT-based metrics. On standard benchmark"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.07316","kind":"arxiv","version":5},"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/2208.07316/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":"2026-07-05T07:27:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FeSN9UWrKZM2v1cEEmAjR9kLzwyzOXdHpcYthlN2QF1UyyLK7lSc00fifhOTBB2nV6EnBunlIZQ1/Gy66MyBCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T22:36:59.868781Z"},"content_sha256":"b5188ac3d4a97e9c63941cfe5a95268cefaef5431893433e2a19a75280cc2594","schema_version":"1.0","event_id":"sha256:b5188ac3d4a97e9c63941cfe5a95268cefaef5431893433e2a19a75280cc2594"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SGFSQFHKD7VMM6VNFJJW3KCCX4/bundle.json","state_url":"https://pith.science/pith/SGFSQFHKD7VMM6VNFJJW3KCCX4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SGFSQFHKD7VMM6VNFJJW3KCCX4/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":"2026-08-04T22:36:59Z","links":{"resolver":"https://pith.science/pith/SGFSQFHKD7VMM6VNFJJW3KCCX4","bundle":"https://pith.science/pith/SGFSQFHKD7VMM6VNFJJW3KCCX4/bundle.json","state":"https://pith.science/pith/SGFSQFHKD7VMM6VNFJJW3KCCX4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SGFSQFHKD7VMM6VNFJJW3KCCX4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:SGFSQFHKD7VMM6VNFJJW3KCCX4","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":"6c79a1a480473c58492f288d5d4db3e1f07f2b88e6abb9d22524712713498a83","cross_cats_sorted":["cs.CR","cs.LG"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2022-08-15T16:30:14Z","title_canon_sha256":"a9a068e80a1948e286cd6fcd4d4b3b47aaa208004b8e99dff968f294033f42ca"},"schema_version":"1.0","source":{"id":"2208.07316","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2208.07316","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"arxiv_version","alias_value":"2208.07316v5","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.07316","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"pith_short_12","alias_value":"SGFSQFHKD7VM","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"pith_short_16","alias_value":"SGFSQFHKD7VMM6VN","created_at":"2026-07-05T07:27:41Z"},{"alias_kind":"pith_short_8","alias_value":"SGFSQFHK","created_at":"2026-07-05T07:27:41Z"}],"graph_snapshots":[{"event_id":"sha256:b5188ac3d4a97e9c63941cfe5a95268cefaef5431893433e2a19a75280cc2594","target":"graph","created_at":"2026-07-05T07:27:41Z","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/2208.07316/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recently proposed BERT-based evaluation metrics for text generation perform well on standard benchmarks but are vulnerable to adversarial attacks, e.g., relating to information correctness. We argue that this stems (in part) from the fact that they are models of semantic similarity. In contrast, we develop evaluation metrics based on Natural Language Inference (NLI), which we deem a more appropriate modeling. We design a preference-based adversarial attack framework and show that our NLI based metrics are much more robust to the attacks than the recent BERT-based metrics. On standard benchmark","authors_text":"Steffen Eger, Yanran Chen","cross_cats":["cs.CR","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2022-08-15T16:30:14Z","title":"MENLI: Robust Evaluation Metrics from Natural Language Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.07316","kind":"arxiv","version":5},"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:bbabc8a2f7f65050b8beebcecff2cc167c86925ff095d99ac082453631e69339","target":"record","created_at":"2026-07-05T07:27:41Z","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":"6c79a1a480473c58492f288d5d4db3e1f07f2b88e6abb9d22524712713498a83","cross_cats_sorted":["cs.CR","cs.LG"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2022-08-15T16:30:14Z","title_canon_sha256":"a9a068e80a1948e286cd6fcd4d4b3b47aaa208004b8e99dff968f294033f42ca"},"schema_version":"1.0","source":{"id":"2208.07316","kind":"arxiv","version":5}},"canonical_sha256":"918b2814ea1feac67aad2a536da842bf146c73e67c8ac4b4697549779c1eb37b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"918b2814ea1feac67aad2a536da842bf146c73e67c8ac4b4697549779c1eb37b","first_computed_at":"2026-07-05T07:27:41.130770Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:27:41.130770Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"UPxOjYu18uKXc4SNS+W6yuSQeEfLJRiuJWc86MTjjZSj/cwM1w2FW7RRoH+hu6IVD0DKpOcGWpbxPMx7yLf9BA==","signature_status":"signed_v1","signed_at":"2026-07-05T07:27:41.131250Z","signed_message":"canonical_sha256_bytes"},"source_id":"2208.07316","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bbabc8a2f7f65050b8beebcecff2cc167c86925ff095d99ac082453631e69339","sha256:b5188ac3d4a97e9c63941cfe5a95268cefaef5431893433e2a19a75280cc2594"],"state_sha256":"863bd9b59950056a79f206f68e45982d0ca5f68ef240e75022981961d777ecb3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"25/PrXHE8X4zw4o9T2vlHmGEXkdUpi8bsHbuZjP62aodJG60ifBNhJSbAmeNQUjmWLOXpmZhw5KQk7rRlccLAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T22:36:59.873951Z","bundle_sha256":"65459d5e179cc8eaa81a43b998b7a8f1b5dfcfb05cdc109197a6972f304e4061"}}