{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:O4KTTK5IAQMP7EMSWM2YOUE7G5","short_pith_number":"pith:O4KTTK5I","canonical_record":{"source":{"id":"2506.07948","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-09T17:11:28Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"d76c58127a814fce192ad2e7e1d38a837476e20240aec18898f322bd59c25772","abstract_canon_sha256":"6f18148e94973a6ce22fc2ad5c26b3d53bb7eacc6c6ed197e4a39872a4d75984"},"schema_version":"1.0"},"canonical_sha256":"771539aba80418ff9192b33587509f37518e9dcc6b4d0e54b19ac780b3735e7f","source":{"kind":"arxiv","id":"2506.07948","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.07948","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"arxiv_version","alias_value":"2506.07948v1","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.07948","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"pith_short_12","alias_value":"O4KTTK5IAQMP","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"pith_short_16","alias_value":"O4KTTK5IAQMP7EMS","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"pith_short_8","alias_value":"O4KTTK5I","created_at":"2026-07-05T11:18:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:O4KTTK5IAQMP7EMSWM2YOUE7G5","target":"record","payload":{"canonical_record":{"source":{"id":"2506.07948","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-09T17:11:28Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"d76c58127a814fce192ad2e7e1d38a837476e20240aec18898f322bd59c25772","abstract_canon_sha256":"6f18148e94973a6ce22fc2ad5c26b3d53bb7eacc6c6ed197e4a39872a4d75984"},"schema_version":"1.0"},"canonical_sha256":"771539aba80418ff9192b33587509f37518e9dcc6b4d0e54b19ac780b3735e7f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:35.900503Z","signature_b64":"/v09ZB56poVonwHJu7R2OS/8uOFtgZUO1wJZ+rQLFaB8IPiBpwReea4ILQwsbLvtXfvLPbjmDv2Ff3WY5fhPDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"771539aba80418ff9192b33587509f37518e9dcc6b4d0e54b19ac780b3735e7f","last_reissued_at":"2026-07-05T11:18:35.899931Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:35.899931Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.07948","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":"2026-07-05T11:18:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8ACcZnTUWDQW8UNfl4u+SJXAyw9JDC1im3S0Tr73aDoU1cfgcT+rIW7BaWTzB8LWZOfQmDU56osF9ZFFZc49Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T23:09:34.488263Z"},"content_sha256":"a3267a3b67ced0123f517d0d9588e98de3cf23a003d7f75cef923b2a6dff9a8b","schema_version":"1.0","event_id":"sha256:a3267a3b67ced0123f517d0d9588e98de3cf23a003d7f75cef923b2a6dff9a8b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:O4KTTK5IAQMP7EMSWM2YOUE7G5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TokenBreak: Bypassing Text Classification Models Through Token Manipulation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Kasimir Schulz, Kenneth Yeung, Kieran Evans","submitted_at":"2025-06-09T17:11:28Z","abstract_excerpt":"Natural Language Processing (NLP) models are used for text-related tasks such as classification and generation. To complete these tasks, input data is first tokenized from human-readable text into a format the model can understand, enabling it to make inferences and understand context. Text classification models can be implemented to guard against threats such as prompt injection attacks against Large Language Models (LLMs), toxic input and cybersecurity risks such as spam emails. In this paper, we introduce TokenBreak: a novel attack that can bypass these protection models by taking advantage"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.07948","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/2506.07948/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":"2026-07-05T11:18:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iszdoCrhxkCvVHHWhvCE4C8q/26HHhjMfe7fKIR76NweDOrsMRAss/nbhMaxQESbU5LdPrR/UjjVinXS9jLTAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T23:09:34.488818Z"},"content_sha256":"9c30cb3ab1b6da30d702fea7a99b6040fa85b6f89d7745409756bd1a7d045624","schema_version":"1.0","event_id":"sha256:9c30cb3ab1b6da30d702fea7a99b6040fa85b6f89d7745409756bd1a7d045624"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/O4KTTK5IAQMP7EMSWM2YOUE7G5/bundle.json","state_url":"https://pith.science/pith/O4KTTK5IAQMP7EMSWM2YOUE7G5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/O4KTTK5IAQMP7EMSWM2YOUE7G5/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":"2026-08-07T23:09:34Z","links":{"resolver":"https://pith.science/pith/O4KTTK5IAQMP7EMSWM2YOUE7G5","bundle":"https://pith.science/pith/O4KTTK5IAQMP7EMSWM2YOUE7G5/bundle.json","state":"https://pith.science/pith/O4KTTK5IAQMP7EMSWM2YOUE7G5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/O4KTTK5IAQMP7EMSWM2YOUE7G5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:O4KTTK5IAQMP7EMSWM2YOUE7G5","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":"6f18148e94973a6ce22fc2ad5c26b3d53bb7eacc6c6ed197e4a39872a4d75984","cross_cats_sorted":["cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-09T17:11:28Z","title_canon_sha256":"d76c58127a814fce192ad2e7e1d38a837476e20240aec18898f322bd59c25772"},"schema_version":"1.0","source":{"id":"2506.07948","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.07948","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"arxiv_version","alias_value":"2506.07948v1","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.07948","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"pith_short_12","alias_value":"O4KTTK5IAQMP","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"pith_short_16","alias_value":"O4KTTK5IAQMP7EMS","created_at":"2026-07-05T11:18:35Z"},{"alias_kind":"pith_short_8","alias_value":"O4KTTK5I","created_at":"2026-07-05T11:18:35Z"}],"graph_snapshots":[{"event_id":"sha256:9c30cb3ab1b6da30d702fea7a99b6040fa85b6f89d7745409756bd1a7d045624","target":"graph","created_at":"2026-07-05T11:18:35Z","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/2506.07948/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Natural Language Processing (NLP) models are used for text-related tasks such as classification and generation. To complete these tasks, input data is first tokenized from human-readable text into a format the model can understand, enabling it to make inferences and understand context. Text classification models can be implemented to guard against threats such as prompt injection attacks against Large Language Models (LLMs), toxic input and cybersecurity risks such as spam emails. In this paper, we introduce TokenBreak: a novel attack that can bypass these protection models by taking advantage","authors_text":"Kasimir Schulz, Kenneth Yeung, Kieran Evans","cross_cats":["cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-09T17:11:28Z","title":"TokenBreak: Bypassing Text Classification Models Through Token Manipulation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.07948","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:a3267a3b67ced0123f517d0d9588e98de3cf23a003d7f75cef923b2a6dff9a8b","target":"record","created_at":"2026-07-05T11:18:35Z","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":"6f18148e94973a6ce22fc2ad5c26b3d53bb7eacc6c6ed197e4a39872a4d75984","cross_cats_sorted":["cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-09T17:11:28Z","title_canon_sha256":"d76c58127a814fce192ad2e7e1d38a837476e20240aec18898f322bd59c25772"},"schema_version":"1.0","source":{"id":"2506.07948","kind":"arxiv","version":1}},"canonical_sha256":"771539aba80418ff9192b33587509f37518e9dcc6b4d0e54b19ac780b3735e7f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"771539aba80418ff9192b33587509f37518e9dcc6b4d0e54b19ac780b3735e7f","first_computed_at":"2026-07-05T11:18:35.899931Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:18:35.899931Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"/v09ZB56poVonwHJu7R2OS/8uOFtgZUO1wJZ+rQLFaB8IPiBpwReea4ILQwsbLvtXfvLPbjmDv2Ff3WY5fhPDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:18:35.900503Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.07948","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a3267a3b67ced0123f517d0d9588e98de3cf23a003d7f75cef923b2a6dff9a8b","sha256:9c30cb3ab1b6da30d702fea7a99b6040fa85b6f89d7745409756bd1a7d045624"],"state_sha256":"533cdbb2040f9fc056b5449a3adf97d04767884b9597e998cca6fa047229eb26"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Q1pXPY9x+rH38iuDekqOEGicHMsMgQk/oeZQgbdO0IG77DczYHVunuoCli5YNK5GhPnikZjJRO4IHMKebqSGCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T23:09:34.492499Z","bundle_sha256":"34ce0ab24fadd67cb1c61cf1113a4966e4c159e337a14e27d175f1bea788e56f"}}