{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:J5ZPVKHFSWZR33Y56CQYD43QRV","short_pith_number":"pith:J5ZPVKHF","canonical_record":{"source":{"id":"2004.04696","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-09T17:26:52Z","cross_cats_sorted":[],"title_canon_sha256":"1255a3a351f709ca33ff001ec1ee0c2d0d807b43ccd20cf72843458d723749a0","abstract_canon_sha256":"3e4d6acea5fa3df82b98e4abcf72dd6166ad1f174d4a9064a21046f5a48c0efe"},"schema_version":"1.0"},"canonical_sha256":"4f72faa8e595b31def1df0a181f3708d77d78c1c70aad13b98d901d29904beed","source":{"kind":"arxiv","id":"2004.04696","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2004.04696","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"arxiv_version","alias_value":"2004.04696v5","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.04696","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"pith_short_12","alias_value":"J5ZPVKHFSWZR","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"pith_short_16","alias_value":"J5ZPVKHFSWZR33Y5","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"pith_short_8","alias_value":"J5ZPVKHF","created_at":"2026-07-05T01:04:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:J5ZPVKHFSWZR33Y56CQYD43QRV","target":"record","payload":{"canonical_record":{"source":{"id":"2004.04696","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-09T17:26:52Z","cross_cats_sorted":[],"title_canon_sha256":"1255a3a351f709ca33ff001ec1ee0c2d0d807b43ccd20cf72843458d723749a0","abstract_canon_sha256":"3e4d6acea5fa3df82b98e4abcf72dd6166ad1f174d4a9064a21046f5a48c0efe"},"schema_version":"1.0"},"canonical_sha256":"4f72faa8e595b31def1df0a181f3708d77d78c1c70aad13b98d901d29904beed","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:04:48.676778Z","signature_b64":"f3lwdo2RvylRaUo36efqQNaRryP/U7XCNRwyMIExJSGfC+IA3tFwfKAcnbEeqNi1selTet/zLLNwWIyB3aHzCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4f72faa8e595b31def1df0a181f3708d77d78c1c70aad13b98d901d29904beed","last_reissued_at":"2026-07-05T01:04:48.676001Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:04:48.676001Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2004.04696","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":"2026-07-05T01:04:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pxnUGNQELkRMaW+eOkM8n9ZKrWvXcqvgLNJFAERurgg+A+VFT8JJdnyN//TswJmy5yj0mG+131tK31nw3SV4CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T11:24:21.997473Z"},"content_sha256":"675cdd5135cd717a821353fa234b2f4d50c345b30c736c4626173882e3e26566","schema_version":"1.0","event_id":"sha256:675cdd5135cd717a821353fa234b2f4d50c345b30c736c4626173882e3e26566"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:J5ZPVKHFSWZR33Y56CQYD43QRV","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"BLEURT: Learning Robust Metrics for Text Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ankur P. Parikh, Dipanjan Das, Thibault Sellam","submitted_at":"2020-04-09T17:26:52Z","abstract_excerpt":"Text generation has made significant advances in the last few years. Yet, evaluation metrics have lagged behind, as the most popular choices (e.g., BLEU and ROUGE) may correlate poorly with human judgments. We propose BLEURT, a learned evaluation metric based on BERT that can model human judgments with a few thousand possibly biased training examples. A key aspect of our approach is a novel pre-training scheme that uses millions of synthetic examples to help the model generalize. BLEURT provides state-of-the-art results on the last three years of the WMT Metrics shared task and the WebNLG Comp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.04696","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/2004.04696/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":"2026-07-05T01:04:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"W89teerK1VBi+sG/8zPDUDbozH+BAk1mLWp3wPUbJt401cuiaq7ggpqnSfMuUlMxBLeDsdsJTS0JK8lUsPdxCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T11:24:21.997979Z"},"content_sha256":"d71d70131ea1a68d3a1fee102428cec4945fc8cd699f860e5bd9f683dce00aab","schema_version":"1.0","event_id":"sha256:d71d70131ea1a68d3a1fee102428cec4945fc8cd699f860e5bd9f683dce00aab"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/J5ZPVKHFSWZR33Y56CQYD43QRV/bundle.json","state_url":"https://pith.science/pith/J5ZPVKHFSWZR33Y56CQYD43QRV/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/J5ZPVKHFSWZR33Y56CQYD43QRV/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":"2026-08-08T11:24:22Z","links":{"resolver":"https://pith.science/pith/J5ZPVKHFSWZR33Y56CQYD43QRV","bundle":"https://pith.science/pith/J5ZPVKHFSWZR33Y56CQYD43QRV/bundle.json","state":"https://pith.science/pith/J5ZPVKHFSWZR33Y56CQYD43QRV/state.json","well_known_bundle":"https://pith.science/.well-known/pith/J5ZPVKHFSWZR33Y56CQYD43QRV/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:J5ZPVKHFSWZR33Y56CQYD43QRV","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":"3e4d6acea5fa3df82b98e4abcf72dd6166ad1f174d4a9064a21046f5a48c0efe","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-09T17:26:52Z","title_canon_sha256":"1255a3a351f709ca33ff001ec1ee0c2d0d807b43ccd20cf72843458d723749a0"},"schema_version":"1.0","source":{"id":"2004.04696","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2004.04696","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"arxiv_version","alias_value":"2004.04696v5","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.04696","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"pith_short_12","alias_value":"J5ZPVKHFSWZR","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"pith_short_16","alias_value":"J5ZPVKHFSWZR33Y5","created_at":"2026-07-05T01:04:48Z"},{"alias_kind":"pith_short_8","alias_value":"J5ZPVKHF","created_at":"2026-07-05T01:04:48Z"}],"graph_snapshots":[{"event_id":"sha256:d71d70131ea1a68d3a1fee102428cec4945fc8cd699f860e5bd9f683dce00aab","target":"graph","created_at":"2026-07-05T01:04:48Z","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/2004.04696/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Text generation has made significant advances in the last few years. Yet, evaluation metrics have lagged behind, as the most popular choices (e.g., BLEU and ROUGE) may correlate poorly with human judgments. We propose BLEURT, a learned evaluation metric based on BERT that can model human judgments with a few thousand possibly biased training examples. A key aspect of our approach is a novel pre-training scheme that uses millions of synthetic examples to help the model generalize. BLEURT provides state-of-the-art results on the last three years of the WMT Metrics shared task and the WebNLG Comp","authors_text":"Ankur P. Parikh, Dipanjan Das, Thibault Sellam","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-09T17:26:52Z","title":"BLEURT: Learning Robust Metrics for Text Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.04696","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":"sha256:675cdd5135cd717a821353fa234b2f4d50c345b30c736c4626173882e3e26566","target":"record","created_at":"2026-07-05T01:04:48Z","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":"3e4d6acea5fa3df82b98e4abcf72dd6166ad1f174d4a9064a21046f5a48c0efe","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-09T17:26:52Z","title_canon_sha256":"1255a3a351f709ca33ff001ec1ee0c2d0d807b43ccd20cf72843458d723749a0"},"schema_version":"1.0","source":{"id":"2004.04696","kind":"arxiv","version":5}},"canonical_sha256":"4f72faa8e595b31def1df0a181f3708d77d78c1c70aad13b98d901d29904beed","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4f72faa8e595b31def1df0a181f3708d77d78c1c70aad13b98d901d29904beed","first_computed_at":"2026-07-05T01:04:48.676001Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T01:04:48.676001Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"f3lwdo2RvylRaUo36efqQNaRryP/U7XCNRwyMIExJSGfC+IA3tFwfKAcnbEeqNi1selTet/zLLNwWIyB3aHzCg==","signature_status":"signed_v1","signed_at":"2026-07-05T01:04:48.676778Z","signed_message":"canonical_sha256_bytes"},"source_id":"2004.04696","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:675cdd5135cd717a821353fa234b2f4d50c345b30c736c4626173882e3e26566","sha256:d71d70131ea1a68d3a1fee102428cec4945fc8cd699f860e5bd9f683dce00aab"],"state_sha256":"24ea61b157343976071b2cba0179f7db13aadcad76df2bed12838eb16445e71f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rbqmLesb2emmWsVQZ8D9k/RbhmekixoKtJvULL8qqAZPzmzjFeRBHa6hzyWQ4IuEOfaSqQNyQOOzr7pRYzxgAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T11:24:22.002501Z","bundle_sha256":"a5c8af2b1a73edf974895191f77fbdce9fa3c7f4b3bc68c5ab5c456cbd0c842a"}}