{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:BLXYOMCW6SR4S54F25Z4JQMRHP","short_pith_number":"pith:BLXYOMCW","canonical_record":{"source":{"id":"2405.05417","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-08T20:37:56Z","cross_cats_sorted":[],"title_canon_sha256":"459f3cf4a99e5b9ada4deedac5af4b53eca8b99102bbe931040ab248dea00f72","abstract_canon_sha256":"80402f28dcef7c44744b3467e7d5a7aa56b0103f5466c3212808f5240d8ff5a1"},"schema_version":"1.0"},"canonical_sha256":"0aef873056f4a3c97785d773c4c1913be3a9a74af797d18a77ac7ad98990f4c7","source":{"kind":"arxiv","id":"2405.05417","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.05417","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"arxiv_version","alias_value":"2405.05417v2","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.05417","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"pith_short_12","alias_value":"BLXYOMCW6SR4","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"pith_short_16","alias_value":"BLXYOMCW6SR4S54F","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"pith_short_8","alias_value":"BLXYOMCW","created_at":"2026-07-05T09:12:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:BLXYOMCW6SR4S54F25Z4JQMRHP","target":"record","payload":{"canonical_record":{"source":{"id":"2405.05417","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-08T20:37:56Z","cross_cats_sorted":[],"title_canon_sha256":"459f3cf4a99e5b9ada4deedac5af4b53eca8b99102bbe931040ab248dea00f72","abstract_canon_sha256":"80402f28dcef7c44744b3467e7d5a7aa56b0103f5466c3212808f5240d8ff5a1"},"schema_version":"1.0"},"canonical_sha256":"0aef873056f4a3c97785d773c4c1913be3a9a74af797d18a77ac7ad98990f4c7","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:12:28.299044Z","signature_b64":"vdb0DfjZBiEUvm5vl/eCqDiNg+Rp5MEHbogDUvv59k4OkxHpV/uMm41eo25Jh3M1BhTN0YRe2HNrihP6hqSLDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0aef873056f4a3c97785d773c4c1913be3a9a74af797d18a77ac7ad98990f4c7","last_reissued_at":"2026-07-05T09:12:28.298504Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:12:28.298504Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.05417","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:12:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PLujSTsrET6l/dNzYDCr9/mgNjELTwORCHDRaNKu+Wv42axvdVH4VV4UY+5JTDhsPvexU76iPxJc5VDPtGnZAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T06:14:22.656656Z"},"content_sha256":"f005f23ca72b579349cef57f9e7d382bf72b8769c1163177f6da165774123b53","schema_version":"1.0","event_id":"sha256:f005f23ca72b579349cef57f9e7d382bf72b8769c1163177f6da165774123b53"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:BLXYOMCW6SR4S54F25Z4JQMRHP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Fishing for Magikarp: Automatically Detecting Under-trained Tokens in Large Language Models","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Max Bartolo, Sander Land","submitted_at":"2024-05-08T20:37:56Z","abstract_excerpt":"The disconnect between tokenizer creation and model training in language models allows for specific inputs, such as the infamous SolidGoldMagikarp token, to induce unwanted model behaviour. Although such `glitch tokens', tokens present in the tokenizer vocabulary but that are nearly or entirely absent during model training, have been observed across various models, a reliable method to identify and address them has been missing. We present a comprehensive analysis of Large Language Model tokenizers, specifically targeting this issue of detecting under-trained tokens. Through a combination of t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.05417","kind":"arxiv","version":2},"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/2405.05417/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":"2026-07-05T09:12:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ltOFV4NvoGGbp/bPi2cMSPUemMIrkPBbbhAP83vPQr/drzEu9DrbhWy1oIS0atk3HsrbU6a33BctQCdNvyDdBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T06:14:22.657178Z"},"content_sha256":"d7cceffb6570c9d92338b3f410066203bf1e4a8d801060b880535b8a282127f1","schema_version":"1.0","event_id":"sha256:d7cceffb6570c9d92338b3f410066203bf1e4a8d801060b880535b8a282127f1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BLXYOMCW6SR4S54F25Z4JQMRHP/bundle.json","state_url":"https://pith.science/pith/BLXYOMCW6SR4S54F25Z4JQMRHP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BLXYOMCW6SR4S54F25Z4JQMRHP/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":"2026-08-09T06:14:22Z","links":{"resolver":"https://pith.science/pith/BLXYOMCW6SR4S54F25Z4JQMRHP","bundle":"https://pith.science/pith/BLXYOMCW6SR4S54F25Z4JQMRHP/bundle.json","state":"https://pith.science/pith/BLXYOMCW6SR4S54F25Z4JQMRHP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BLXYOMCW6SR4S54F25Z4JQMRHP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:BLXYOMCW6SR4S54F25Z4JQMRHP","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":"80402f28dcef7c44744b3467e7d5a7aa56b0103f5466c3212808f5240d8ff5a1","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-08T20:37:56Z","title_canon_sha256":"459f3cf4a99e5b9ada4deedac5af4b53eca8b99102bbe931040ab248dea00f72"},"schema_version":"1.0","source":{"id":"2405.05417","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.05417","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"arxiv_version","alias_value":"2405.05417v2","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.05417","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"pith_short_12","alias_value":"BLXYOMCW6SR4","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"pith_short_16","alias_value":"BLXYOMCW6SR4S54F","created_at":"2026-07-05T09:12:28Z"},{"alias_kind":"pith_short_8","alias_value":"BLXYOMCW","created_at":"2026-07-05T09:12:28Z"}],"graph_snapshots":[{"event_id":"sha256:d7cceffb6570c9d92338b3f410066203bf1e4a8d801060b880535b8a282127f1","target":"graph","created_at":"2026-07-05T09:12:28Z","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/2405.05417/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The disconnect between tokenizer creation and model training in language models allows for specific inputs, such as the infamous SolidGoldMagikarp token, to induce unwanted model behaviour. Although such `glitch tokens', tokens present in the tokenizer vocabulary but that are nearly or entirely absent during model training, have been observed across various models, a reliable method to identify and address them has been missing. We present a comprehensive analysis of Large Language Model tokenizers, specifically targeting this issue of detecting under-trained tokens. Through a combination of t","authors_text":"Max Bartolo, Sander Land","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-08T20:37:56Z","title":"Fishing for Magikarp: Automatically Detecting Under-trained Tokens in Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.05417","kind":"arxiv","version":2},"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:f005f23ca72b579349cef57f9e7d382bf72b8769c1163177f6da165774123b53","target":"record","created_at":"2026-07-05T09:12:28Z","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":"80402f28dcef7c44744b3467e7d5a7aa56b0103f5466c3212808f5240d8ff5a1","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-08T20:37:56Z","title_canon_sha256":"459f3cf4a99e5b9ada4deedac5af4b53eca8b99102bbe931040ab248dea00f72"},"schema_version":"1.0","source":{"id":"2405.05417","kind":"arxiv","version":2}},"canonical_sha256":"0aef873056f4a3c97785d773c4c1913be3a9a74af797d18a77ac7ad98990f4c7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0aef873056f4a3c97785d773c4c1913be3a9a74af797d18a77ac7ad98990f4c7","first_computed_at":"2026-07-05T09:12:28.298504Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:12:28.298504Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vdb0DfjZBiEUvm5vl/eCqDiNg+Rp5MEHbogDUvv59k4OkxHpV/uMm41eo25Jh3M1BhTN0YRe2HNrihP6hqSLDw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:12:28.299044Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.05417","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f005f23ca72b579349cef57f9e7d382bf72b8769c1163177f6da165774123b53","sha256:d7cceffb6570c9d92338b3f410066203bf1e4a8d801060b880535b8a282127f1"],"state_sha256":"d30bd4d25f538d1f6dd4b183963b155f6e891f1c0d23f05f347ad016de770fb7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CSPchyLHQoYbBlfjO+7ar5HhYAtKhX1yNRGn50DqDJIkdrxEwZ7RG65BtcaJTPx7jYfosN3QFTLeVcddeWxyDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T06:14:22.660744Z","bundle_sha256":"e829bbe5c4b7e371daf2c331ff35514287918f50d8c60029781a6248bed6d76d"}}