{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:JNV762FF465NAIEVGYVLRTNBGI","short_pith_number":"pith:JNV762FF","canonical_record":{"source":{"id":"2005.00456","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-05-01T15:50:50Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4c33347dbc878f8659d6d1f20803ac24863d7231716f63634284dca2380e34d1","abstract_canon_sha256":"000487c4904bc4fe61bbd92669cf09ba3bff3a873f4ea9b90350e739c9581037"},"schema_version":"1.0"},"canonical_sha256":"4b6bff68a5e7bad02095362ab8cda132051cf6d28c67dffaafe0e73416dc1b0e","source":{"kind":"arxiv","id":"2005.00456","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2005.00456","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"arxiv_version","alias_value":"2005.00456v1","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2005.00456","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"pith_short_12","alias_value":"JNV762FF465N","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"pith_short_16","alias_value":"JNV762FF465NAIEV","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"pith_short_8","alias_value":"JNV762FF","created_at":"2026-07-05T00:59:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:JNV762FF465NAIEVGYVLRTNBGI","target":"record","payload":{"canonical_record":{"source":{"id":"2005.00456","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-05-01T15:50:50Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"4c33347dbc878f8659d6d1f20803ac24863d7231716f63634284dca2380e34d1","abstract_canon_sha256":"000487c4904bc4fe61bbd92669cf09ba3bff3a873f4ea9b90350e739c9581037"},"schema_version":"1.0"},"canonical_sha256":"4b6bff68a5e7bad02095362ab8cda132051cf6d28c67dffaafe0e73416dc1b0e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:59:41.163272Z","signature_b64":"rZxQB8/bYHVZIc1ijIfS9vzz6D7zoR8RRB2iGR2RnHyPgsoFOVq74vN+GZL+0z8W12H4gHXAdStcHYafpeM7Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4b6bff68a5e7bad02095362ab8cda132051cf6d28c67dffaafe0e73416dc1b0e","last_reissued_at":"2026-07-05T00:59:41.162803Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:59:41.162803Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2005.00456","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":"2026-07-05T00:59:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kFBAtIl1EJYTK7JPpUSLumyKnVjmL5JThsTMW5fDcV5Ojb2oqQieM7emX2yWk1TMlcfld35FS6RY3FD1Ykk6CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T10:41:35.646585Z"},"content_sha256":"3cabe95a7e200a06c2c4bd5f494867e26947bd964836d2e3a9f2eed4a94ed19d","schema_version":"1.0","event_id":"sha256:3cabe95a7e200a06c2c4bd5f494867e26947bd964836d2e3a9f2eed4a94ed19d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:JNV762FF465NAIEVGYVLRTNBGI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"USR: An Unsupervised and Reference Free Evaluation Metric for Dialog Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Maxine Eskenazi, Shikib Mehri","submitted_at":"2020-05-01T15:50:50Z","abstract_excerpt":"The lack of meaningful automatic evaluation metrics for dialog has impeded open-domain dialog research. Standard language generation metrics have been shown to be ineffective for evaluating dialog models. To this end, this paper presents USR, an UnSupervised and Reference-free evaluation metric for dialog. USR is a reference-free metric that trains unsupervised models to measure several desirable qualities of dialog. USR is shown to strongly correlate with human judgment on both Topical-Chat (turn-level: 0.42, system-level: 1.0) and PersonaChat (turn-level: 0.48 and system-level: 1.0). USR add"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2005.00456","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2005.00456/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":"2026-07-05T00:59:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RjKCycgOdvhXOcmjNtyVe4wTQGw5snumxXsCIDjwja/jF28MrCezhNZYSOKqouGml10e7TBR6yG64uFMrY+xBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T10:41:35.655284Z"},"content_sha256":"15b2f0df7f70b0ddea5e5d739b652d730836de54eafd19a357a9917a0bfb9254","schema_version":"1.0","event_id":"sha256:15b2f0df7f70b0ddea5e5d739b652d730836de54eafd19a357a9917a0bfb9254"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JNV762FF465NAIEVGYVLRTNBGI/bundle.json","state_url":"https://pith.science/pith/JNV762FF465NAIEVGYVLRTNBGI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JNV762FF465NAIEVGYVLRTNBGI/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":"2026-08-11T10:41:35Z","links":{"resolver":"https://pith.science/pith/JNV762FF465NAIEVGYVLRTNBGI","bundle":"https://pith.science/pith/JNV762FF465NAIEVGYVLRTNBGI/bundle.json","state":"https://pith.science/pith/JNV762FF465NAIEVGYVLRTNBGI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JNV762FF465NAIEVGYVLRTNBGI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:JNV762FF465NAIEVGYVLRTNBGI","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":"000487c4904bc4fe61bbd92669cf09ba3bff3a873f4ea9b90350e739c9581037","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-05-01T15:50:50Z","title_canon_sha256":"4c33347dbc878f8659d6d1f20803ac24863d7231716f63634284dca2380e34d1"},"schema_version":"1.0","source":{"id":"2005.00456","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2005.00456","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"arxiv_version","alias_value":"2005.00456v1","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2005.00456","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"pith_short_12","alias_value":"JNV762FF465N","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"pith_short_16","alias_value":"JNV762FF465NAIEV","created_at":"2026-07-05T00:59:41Z"},{"alias_kind":"pith_short_8","alias_value":"JNV762FF","created_at":"2026-07-05T00:59:41Z"}],"graph_snapshots":[{"event_id":"sha256:15b2f0df7f70b0ddea5e5d739b652d730836de54eafd19a357a9917a0bfb9254","target":"graph","created_at":"2026-07-05T00:59:41Z","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/2005.00456/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The lack of meaningful automatic evaluation metrics for dialog has impeded open-domain dialog research. Standard language generation metrics have been shown to be ineffective for evaluating dialog models. To this end, this paper presents USR, an UnSupervised and Reference-free evaluation metric for dialog. USR is a reference-free metric that trains unsupervised models to measure several desirable qualities of dialog. USR is shown to strongly correlate with human judgment on both Topical-Chat (turn-level: 0.42, system-level: 1.0) and PersonaChat (turn-level: 0.48 and system-level: 1.0). USR add","authors_text":"Maxine Eskenazi, Shikib Mehri","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-05-01T15:50:50Z","title":"USR: An Unsupervised and Reference Free Evaluation Metric for Dialog Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2005.00456","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":"sha256:3cabe95a7e200a06c2c4bd5f494867e26947bd964836d2e3a9f2eed4a94ed19d","target":"record","created_at":"2026-07-05T00:59:41Z","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":"000487c4904bc4fe61bbd92669cf09ba3bff3a873f4ea9b90350e739c9581037","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-05-01T15:50:50Z","title_canon_sha256":"4c33347dbc878f8659d6d1f20803ac24863d7231716f63634284dca2380e34d1"},"schema_version":"1.0","source":{"id":"2005.00456","kind":"arxiv","version":1}},"canonical_sha256":"4b6bff68a5e7bad02095362ab8cda132051cf6d28c67dffaafe0e73416dc1b0e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4b6bff68a5e7bad02095362ab8cda132051cf6d28c67dffaafe0e73416dc1b0e","first_computed_at":"2026-07-05T00:59:41.162803Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T00:59:41.162803Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rZxQB8/bYHVZIc1ijIfS9vzz6D7zoR8RRB2iGR2RnHyPgsoFOVq74vN+GZL+0z8W12H4gHXAdStcHYafpeM7Ag==","signature_status":"signed_v1","signed_at":"2026-07-05T00:59:41.163272Z","signed_message":"canonical_sha256_bytes"},"source_id":"2005.00456","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3cabe95a7e200a06c2c4bd5f494867e26947bd964836d2e3a9f2eed4a94ed19d","sha256:15b2f0df7f70b0ddea5e5d739b652d730836de54eafd19a357a9917a0bfb9254"],"state_sha256":"092162f4812643f48896188a9df966fdd77b9ee024a7e6a073bda2e041a8016e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P0FDBcPvTz0B+GOQSXK4DX8rFrQv8Yh7YVY4nJKYT9FdAaKgIeUMLoCm/2IPchBzY32fJ58h7gJUKLtJIuXvBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-11T10:41:35.673252Z","bundle_sha256":"e3758ea50d781f2f6ce51fad5ac6eab0c64d70fac9ae3177a1336a5b6211be34"}}