{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:XKJVAGM2B5DNBPRCOKAD2D4HN6","short_pith_number":"pith:XKJVAGM2","canonical_record":{"source":{"id":"2205.05092","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-10T18:00:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1978b2ac9abf229c04dccb10969d594ad0a8a78a0c56f72f87d03e2d8d557ba7","abstract_canon_sha256":"d6e8e48a62a3bc09983c0dd986ba7cba24c6518a01536403e5958b2b9a7f8619"},"schema_version":"1.0"},"canonical_sha256":"ba9350199a0f46d0be2272803d0f876fbac22d72e96a9307d1426b238a636af1","source":{"kind":"arxiv","id":"2205.05092","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2205.05092","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"arxiv_version","alias_value":"2205.05092v1","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.05092","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"pith_short_12","alias_value":"XKJVAGM2B5DN","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"pith_short_16","alias_value":"XKJVAGM2B5DNBPRC","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"pith_short_8","alias_value":"XKJVAGM2","created_at":"2026-07-05T04:22:22Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:XKJVAGM2B5DNBPRCOKAD2D4HN6","target":"record","payload":{"canonical_record":{"source":{"id":"2205.05092","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-10T18:00:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1978b2ac9abf229c04dccb10969d594ad0a8a78a0c56f72f87d03e2d8d557ba7","abstract_canon_sha256":"d6e8e48a62a3bc09983c0dd986ba7cba24c6518a01536403e5958b2b9a7f8619"},"schema_version":"1.0"},"canonical_sha256":"ba9350199a0f46d0be2272803d0f876fbac22d72e96a9307d1426b238a636af1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:22:22.322508Z","signature_b64":"PRX2baqhxjnCmKQuuIUOxVxNA9YQdR80zAH+/iYI71OEPi0nwccLWlny6ogEQdW9jjlmCVxGiqs4Bs90RvTLAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba9350199a0f46d0be2272803d0f876fbac22d72e96a9307d1426b238a636af1","last_reissued_at":"2026-07-05T04:22:22.322068Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:22:22.322068Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2205.05092","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":"2026-07-05T04:22:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4zZYa2Ih0q8WkNCwUdC3EruatCv8KK+ss+JNAm/GGem1XTW67/vfzo8/rjUs7WLslF7TV/XM1TH8UIHteY1tDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:12:52.793917Z"},"content_sha256":"5959deedcb35e233a1fe370ffda1b9bcbc699715b2af2b9bd78db5a1b74e6240","schema_version":"1.0","event_id":"sha256:5959deedcb35e233a1fe370ffda1b9bcbc699715b2af2b9bd78db5a1b74e6240"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:XKJVAGM2B5DNBPRCOKAD2D4HN6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Problems with Cosine as a Measure of Embedding Similarity for High Frequency Words","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Dallas Card, Dan Jurafsky, Kaitlyn Zhou, Kawin Ethayarajh","submitted_at":"2022-05-10T18:00:06Z","abstract_excerpt":"Cosine similarity of contextual embeddings is used in many NLP tasks (e.g., QA, IR, MT) and metrics (e.g., BERTScore). Here, we uncover systematic ways in which word similarities estimated by cosine over BERT embeddings are understated and trace this effect to training data frequency. We find that relative to human judgements, cosine similarity underestimates the similarity of frequent words with other instances of the same word or other words across contexts, even after controlling for polysemy and other factors. We conjecture that this underestimation of similarity for high frequency words i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.05092","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/2205.05092/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":"2026-07-05T04:22:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DlVKl8JDeOE/WDyarh261X73BrJq5+fibJiBZyVi94Yi9qe5rt7q29nh5qA+eOPIBBdTpMnGKrExM+eVPTsNDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:12:52.794637Z"},"content_sha256":"fee66d628ca9a02034c589312566d8e37a11aefcd1dedc17b0bce8b4c62c292b","schema_version":"1.0","event_id":"sha256:fee66d628ca9a02034c589312566d8e37a11aefcd1dedc17b0bce8b4c62c292b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XKJVAGM2B5DNBPRCOKAD2D4HN6/bundle.json","state_url":"https://pith.science/pith/XKJVAGM2B5DNBPRCOKAD2D4HN6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XKJVAGM2B5DNBPRCOKAD2D4HN6/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":"2026-08-09T00:12:52Z","links":{"resolver":"https://pith.science/pith/XKJVAGM2B5DNBPRCOKAD2D4HN6","bundle":"https://pith.science/pith/XKJVAGM2B5DNBPRCOKAD2D4HN6/bundle.json","state":"https://pith.science/pith/XKJVAGM2B5DNBPRCOKAD2D4HN6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XKJVAGM2B5DNBPRCOKAD2D4HN6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:XKJVAGM2B5DNBPRCOKAD2D4HN6","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":"d6e8e48a62a3bc09983c0dd986ba7cba24c6518a01536403e5958b2b9a7f8619","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-10T18:00:06Z","title_canon_sha256":"1978b2ac9abf229c04dccb10969d594ad0a8a78a0c56f72f87d03e2d8d557ba7"},"schema_version":"1.0","source":{"id":"2205.05092","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2205.05092","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"arxiv_version","alias_value":"2205.05092v1","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.05092","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"pith_short_12","alias_value":"XKJVAGM2B5DN","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"pith_short_16","alias_value":"XKJVAGM2B5DNBPRC","created_at":"2026-07-05T04:22:22Z"},{"alias_kind":"pith_short_8","alias_value":"XKJVAGM2","created_at":"2026-07-05T04:22:22Z"}],"graph_snapshots":[{"event_id":"sha256:fee66d628ca9a02034c589312566d8e37a11aefcd1dedc17b0bce8b4c62c292b","target":"graph","created_at":"2026-07-05T04:22:22Z","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/2205.05092/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Cosine similarity of contextual embeddings is used in many NLP tasks (e.g., QA, IR, MT) and metrics (e.g., BERTScore). Here, we uncover systematic ways in which word similarities estimated by cosine over BERT embeddings are understated and trace this effect to training data frequency. We find that relative to human judgements, cosine similarity underestimates the similarity of frequent words with other instances of the same word or other words across contexts, even after controlling for polysemy and other factors. We conjecture that this underestimation of similarity for high frequency words i","authors_text":"Dallas Card, Dan Jurafsky, Kaitlyn Zhou, Kawin Ethayarajh","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-10T18:00:06Z","title":"Problems with Cosine as a Measure of Embedding Similarity for High Frequency Words"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.05092","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":"sha256:5959deedcb35e233a1fe370ffda1b9bcbc699715b2af2b9bd78db5a1b74e6240","target":"record","created_at":"2026-07-05T04:22:22Z","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":"d6e8e48a62a3bc09983c0dd986ba7cba24c6518a01536403e5958b2b9a7f8619","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-10T18:00:06Z","title_canon_sha256":"1978b2ac9abf229c04dccb10969d594ad0a8a78a0c56f72f87d03e2d8d557ba7"},"schema_version":"1.0","source":{"id":"2205.05092","kind":"arxiv","version":1}},"canonical_sha256":"ba9350199a0f46d0be2272803d0f876fbac22d72e96a9307d1426b238a636af1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ba9350199a0f46d0be2272803d0f876fbac22d72e96a9307d1426b238a636af1","first_computed_at":"2026-07-05T04:22:22.322068Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:22:22.322068Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"PRX2baqhxjnCmKQuuIUOxVxNA9YQdR80zAH+/iYI71OEPi0nwccLWlny6ogEQdW9jjlmCVxGiqs4Bs90RvTLAw==","signature_status":"signed_v1","signed_at":"2026-07-05T04:22:22.322508Z","signed_message":"canonical_sha256_bytes"},"source_id":"2205.05092","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5959deedcb35e233a1fe370ffda1b9bcbc699715b2af2b9bd78db5a1b74e6240","sha256:fee66d628ca9a02034c589312566d8e37a11aefcd1dedc17b0bce8b4c62c292b"],"state_sha256":"ae8023d23b25bac8024dd714d5cb291f693e6b9afce799b3d2ed1633b7da03a3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xhvzafoUmjWSpwB05NT1bJzz3QRD73hu9/CSKkPBmQlxyn1zQ72w22K8/ShGFb8m/4F1BF3vH8ZGDansGL+8Ag==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T00:12:52.801251Z","bundle_sha256":"45f733fb61a43e9fde8ef51aa3f08ae2dfce508e14ec54fdd6f94e630881dc07"}}