{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:ULBQLMH3GPUDHOLMYLKAWVWU6D","short_pith_number":"pith:ULBQLMH3","canonical_record":{"source":{"id":"2206.02608","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-06-06T13:27:26Z","cross_cats_sorted":[],"title_canon_sha256":"a0b627cfbd43fe73d3ee8eb64d43c03d8afdf01c3eb532b1457c15ccd8e49676","abstract_canon_sha256":"828b41d62da5a677952996231cc2def3f07eeee41e2feee127b5a8e51e732de8"},"schema_version":"1.0"},"canonical_sha256":"a2c305b0fb33e833b96cc2d40b56d4f0eea99af9816e3b83a636931d6f24aa62","source":{"kind":"arxiv","id":"2206.02608","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2206.02608","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"arxiv_version","alias_value":"2206.02608v1","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.02608","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"pith_short_12","alias_value":"ULBQLMH3GPUD","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"pith_short_16","alias_value":"ULBQLMH3GPUDHOLM","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"pith_short_8","alias_value":"ULBQLMH3","created_at":"2026-07-05T04:29:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:ULBQLMH3GPUDHOLMYLKAWVWU6D","target":"record","payload":{"canonical_record":{"source":{"id":"2206.02608","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-06-06T13:27:26Z","cross_cats_sorted":[],"title_canon_sha256":"a0b627cfbd43fe73d3ee8eb64d43c03d8afdf01c3eb532b1457c15ccd8e49676","abstract_canon_sha256":"828b41d62da5a677952996231cc2def3f07eeee41e2feee127b5a8e51e732de8"},"schema_version":"1.0"},"canonical_sha256":"a2c305b0fb33e833b96cc2d40b56d4f0eea99af9816e3b83a636931d6f24aa62","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:29:19.565530Z","signature_b64":"9DTl7EXk0HzVTlxxAwvEKL4y3AiBoElSCdr97Tbh1xjut70FIl0DR6EG3WgQw85q1CQv+7d4Qk+RLoRapyfMDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a2c305b0fb33e833b96cc2d40b56d4f0eea99af9816e3b83a636931d6f24aa62","last_reissued_at":"2026-07-05T04:29:19.565044Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:29:19.565044Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2206.02608","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":"2026-07-05T04:29:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0V7/0Gt7w960+cu/8/Ehvegtu1a82PsLkUBU52CLgKiFjKuA1ZTmAanlEVLhBmrLyP3TGRSupKL0xnMtQXSYDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T02:23:50.828369Z"},"content_sha256":"4f0a5cb17f57b067c90e67f304708d817b78a1e6a8810f06205a4850bedcaf5c","schema_version":"1.0","event_id":"sha256:4f0a5cb17f57b067c90e67f304708d817b78a1e6a8810f06205a4850bedcaf5c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:ULBQLMH3GPUDHOLMYLKAWVWU6D","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"What do tokens know about their characters and how do they know it?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ayush Kaushal, Kyle Mahowald","submitted_at":"2022-06-06T13:27:26Z","abstract_excerpt":"Pre-trained language models (PLMs) that use subword tokenization schemes can succeed at a variety of language tasks that require character-level information, despite lacking explicit access to the character composition of tokens. Here, studying a range of models (e.g., GPT- J, BERT, RoBERTa, GloVe), we probe what word pieces encode about character-level information by training classifiers to predict the presence or absence of a particular alphabetical character in a token, based on its embedding (e.g., probing whether the model embedding for \"cat\" encodes that it contains the character \"a\"). W"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.02608","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/2206.02608/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":"2026-07-05T04:29:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qFsu6sknadOMDVk2AttRJ2r9n8uinHQcbHA052JFQ6J3cJvwutkhEJxgqac18o+qBH8gGWio9gm66hi+keubBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T02:23:50.828854Z"},"content_sha256":"785725e857dfca13f2c6083387cde0237c448d06a4b789b79431f66719d93c4f","schema_version":"1.0","event_id":"sha256:785725e857dfca13f2c6083387cde0237c448d06a4b789b79431f66719d93c4f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ULBQLMH3GPUDHOLMYLKAWVWU6D/bundle.json","state_url":"https://pith.science/pith/ULBQLMH3GPUDHOLMYLKAWVWU6D/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ULBQLMH3GPUDHOLMYLKAWVWU6D/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":"2026-08-08T02:23:50Z","links":{"resolver":"https://pith.science/pith/ULBQLMH3GPUDHOLMYLKAWVWU6D","bundle":"https://pith.science/pith/ULBQLMH3GPUDHOLMYLKAWVWU6D/bundle.json","state":"https://pith.science/pith/ULBQLMH3GPUDHOLMYLKAWVWU6D/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ULBQLMH3GPUDHOLMYLKAWVWU6D/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:ULBQLMH3GPUDHOLMYLKAWVWU6D","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":"828b41d62da5a677952996231cc2def3f07eeee41e2feee127b5a8e51e732de8","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-06-06T13:27:26Z","title_canon_sha256":"a0b627cfbd43fe73d3ee8eb64d43c03d8afdf01c3eb532b1457c15ccd8e49676"},"schema_version":"1.0","source":{"id":"2206.02608","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2206.02608","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"arxiv_version","alias_value":"2206.02608v1","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.02608","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"pith_short_12","alias_value":"ULBQLMH3GPUD","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"pith_short_16","alias_value":"ULBQLMH3GPUDHOLM","created_at":"2026-07-05T04:29:19Z"},{"alias_kind":"pith_short_8","alias_value":"ULBQLMH3","created_at":"2026-07-05T04:29:19Z"}],"graph_snapshots":[{"event_id":"sha256:785725e857dfca13f2c6083387cde0237c448d06a4b789b79431f66719d93c4f","target":"graph","created_at":"2026-07-05T04:29:19Z","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/2206.02608/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Pre-trained language models (PLMs) that use subword tokenization schemes can succeed at a variety of language tasks that require character-level information, despite lacking explicit access to the character composition of tokens. Here, studying a range of models (e.g., GPT- J, BERT, RoBERTa, GloVe), we probe what word pieces encode about character-level information by training classifiers to predict the presence or absence of a particular alphabetical character in a token, based on its embedding (e.g., probing whether the model embedding for \"cat\" encodes that it contains the character \"a\"). W","authors_text":"Ayush Kaushal, Kyle Mahowald","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-06-06T13:27:26Z","title":"What do tokens know about their characters and how do they know it?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.02608","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":"sha256:4f0a5cb17f57b067c90e67f304708d817b78a1e6a8810f06205a4850bedcaf5c","target":"record","created_at":"2026-07-05T04:29:19Z","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":"828b41d62da5a677952996231cc2def3f07eeee41e2feee127b5a8e51e732de8","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-06-06T13:27:26Z","title_canon_sha256":"a0b627cfbd43fe73d3ee8eb64d43c03d8afdf01c3eb532b1457c15ccd8e49676"},"schema_version":"1.0","source":{"id":"2206.02608","kind":"arxiv","version":1}},"canonical_sha256":"a2c305b0fb33e833b96cc2d40b56d4f0eea99af9816e3b83a636931d6f24aa62","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a2c305b0fb33e833b96cc2d40b56d4f0eea99af9816e3b83a636931d6f24aa62","first_computed_at":"2026-07-05T04:29:19.565044Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:29:19.565044Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"9DTl7EXk0HzVTlxxAwvEKL4y3AiBoElSCdr97Tbh1xjut70FIl0DR6EG3WgQw85q1CQv+7d4Qk+RLoRapyfMDg==","signature_status":"signed_v1","signed_at":"2026-07-05T04:29:19.565530Z","signed_message":"canonical_sha256_bytes"},"source_id":"2206.02608","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4f0a5cb17f57b067c90e67f304708d817b78a1e6a8810f06205a4850bedcaf5c","sha256:785725e857dfca13f2c6083387cde0237c448d06a4b789b79431f66719d93c4f"],"state_sha256":"e0c413e38123366295cf774e6cca9266c9ea0438e1ab61e5bb0c78dfae90bf6d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4GWzzLKBgeMcc+paGWGUTJ6yJDZQ14c23xDclgoBbVRNn30C6gYP7sAtpUFS1o3c8C2jlJorA6JlnlrgoI7OCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T02:23:50.832172Z","bundle_sha256":"8dffcdce7bfe642ee9f66c26134ef2882edc932432400e97649bfaf6b10890be"}}