{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3L5A23ND4YVFJGEWGE3G3VQTQK","short_pith_number":"pith:3L5A23ND","canonical_record":{"source":{"id":"2504.02122","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-02T20:50:31Z","cross_cats_sorted":[],"title_canon_sha256":"43ed20d71851e3195d9a5bcb20726976d86f23040fb0f0dcdf0c25bbefe89784","abstract_canon_sha256":"4c1e9c20d70bc2aec8e7f4613c94de9dcf6fbd19970716bd68166f02982eddd9"},"schema_version":"1.0"},"canonical_sha256":"dafa0d6da3e62a54989631366dd613828f0b8c91ae3a2029398ab051f391879d","source":{"kind":"arxiv","id":"2504.02122","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.02122","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"arxiv_version","alias_value":"2504.02122v2","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.02122","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"pith_short_12","alias_value":"3L5A23ND4YVF","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"pith_short_16","alias_value":"3L5A23ND4YVFJGEW","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"pith_short_8","alias_value":"3L5A23ND","created_at":"2026-07-05T11:51:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3L5A23ND4YVFJGEWGE3G3VQTQK","target":"record","payload":{"canonical_record":{"source":{"id":"2504.02122","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-02T20:50:31Z","cross_cats_sorted":[],"title_canon_sha256":"43ed20d71851e3195d9a5bcb20726976d86f23040fb0f0dcdf0c25bbefe89784","abstract_canon_sha256":"4c1e9c20d70bc2aec8e7f4613c94de9dcf6fbd19970716bd68166f02982eddd9"},"schema_version":"1.0"},"canonical_sha256":"dafa0d6da3e62a54989631366dd613828f0b8c91ae3a2029398ab051f391879d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:51:24.836069Z","signature_b64":"s/Go/AG6ZBCCWEv4Dx5VYdJsyIaITwAs8gUIexUr4oNlbI/mEbtjZVvdmbqN0PdRG+fvkNi9flvHnENs5szODQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dafa0d6da3e62a54989631366dd613828f0b8c91ae3a2029398ab051f391879d","last_reissued_at":"2026-07-05T11:51:24.835552Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:51:24.835552Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2504.02122","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":"2026-07-05T11:51:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DBeukGyzj8A22TANUXoH8JXWiTKvMU4wL9pr2o0dmevbyiyCVOAK0rWHG2v+GA41hfoYCjadbI/TFAVeGf5XCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T01:53:07.494080Z"},"content_sha256":"cf0faa4f4abfe450f5d78fe9a3db1ee8b50f78811ce64973f75383f3c6f2dad7","schema_version":"1.0","event_id":"sha256:cf0faa4f4abfe450f5d78fe9a3db1ee8b50f78811ce64973f75383f3c6f2dad7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3L5A23ND4YVFJGEWGE3G3VQTQK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Overcoming Vocabulary Constraints with Pixel-level Fallback","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hendra Setiawan, Jonas F. Lotz, Stephan Peitz, Yova Kementchedjhieva","submitted_at":"2025-04-02T20:50:31Z","abstract_excerpt":"Subword tokenization requires balancing computational efficiency and vocabulary coverage, which often leads to suboptimal performance on languages and scripts not prioritized during training. We propose to augment pretrained language models with a vocabulary-free encoder that generates input embeddings from text rendered as pixels. Through experiments on English-centric language models, we demonstrate that our approach substantially improves machine translation performance and facilitates effective cross-lingual transfer, outperforming tokenizer-based methods. Furthermore, we find that pixel-b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.02122","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/2504.02122/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":"2026-07-05T11:51:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RbKUnxIQ0ovBn3Jhh+Adsjd9HMJwQgsUd5H3tUty8O/htklcRFIBD8hIhjYX4XFWh8lliMK1l3XxQ93wSSk4BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T01:53:07.495103Z"},"content_sha256":"d49696f263d5a4aefff30459fcab47e8b67da4b85c3cf02308fc8183642c9683","schema_version":"1.0","event_id":"sha256:d49696f263d5a4aefff30459fcab47e8b67da4b85c3cf02308fc8183642c9683"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3L5A23ND4YVFJGEWGE3G3VQTQK/bundle.json","state_url":"https://pith.science/pith/3L5A23ND4YVFJGEWGE3G3VQTQK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3L5A23ND4YVFJGEWGE3G3VQTQK/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":"2026-08-15T01:53:07Z","links":{"resolver":"https://pith.science/pith/3L5A23ND4YVFJGEWGE3G3VQTQK","bundle":"https://pith.science/pith/3L5A23ND4YVFJGEWGE3G3VQTQK/bundle.json","state":"https://pith.science/pith/3L5A23ND4YVFJGEWGE3G3VQTQK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3L5A23ND4YVFJGEWGE3G3VQTQK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3L5A23ND4YVFJGEWGE3G3VQTQK","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":"4c1e9c20d70bc2aec8e7f4613c94de9dcf6fbd19970716bd68166f02982eddd9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-02T20:50:31Z","title_canon_sha256":"43ed20d71851e3195d9a5bcb20726976d86f23040fb0f0dcdf0c25bbefe89784"},"schema_version":"1.0","source":{"id":"2504.02122","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.02122","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"arxiv_version","alias_value":"2504.02122v2","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.02122","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"pith_short_12","alias_value":"3L5A23ND4YVF","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"pith_short_16","alias_value":"3L5A23ND4YVFJGEW","created_at":"2026-07-05T11:51:24Z"},{"alias_kind":"pith_short_8","alias_value":"3L5A23ND","created_at":"2026-07-05T11:51:24Z"}],"graph_snapshots":[{"event_id":"sha256:d49696f263d5a4aefff30459fcab47e8b67da4b85c3cf02308fc8183642c9683","target":"graph","created_at":"2026-07-05T11:51:24Z","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/2504.02122/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Subword tokenization requires balancing computational efficiency and vocabulary coverage, which often leads to suboptimal performance on languages and scripts not prioritized during training. We propose to augment pretrained language models with a vocabulary-free encoder that generates input embeddings from text rendered as pixels. Through experiments on English-centric language models, we demonstrate that our approach substantially improves machine translation performance and facilitates effective cross-lingual transfer, outperforming tokenizer-based methods. Furthermore, we find that pixel-b","authors_text":"Hendra Setiawan, Jonas F. Lotz, Stephan Peitz, Yova Kementchedjhieva","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-02T20:50:31Z","title":"Overcoming Vocabulary Constraints with Pixel-level Fallback"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.02122","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:cf0faa4f4abfe450f5d78fe9a3db1ee8b50f78811ce64973f75383f3c6f2dad7","target":"record","created_at":"2026-07-05T11:51:24Z","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":"4c1e9c20d70bc2aec8e7f4613c94de9dcf6fbd19970716bd68166f02982eddd9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-02T20:50:31Z","title_canon_sha256":"43ed20d71851e3195d9a5bcb20726976d86f23040fb0f0dcdf0c25bbefe89784"},"schema_version":"1.0","source":{"id":"2504.02122","kind":"arxiv","version":2}},"canonical_sha256":"dafa0d6da3e62a54989631366dd613828f0b8c91ae3a2029398ab051f391879d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dafa0d6da3e62a54989631366dd613828f0b8c91ae3a2029398ab051f391879d","first_computed_at":"2026-07-05T11:51:24.835552Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:51:24.835552Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"s/Go/AG6ZBCCWEv4Dx5VYdJsyIaITwAs8gUIexUr4oNlbI/mEbtjZVvdmbqN0PdRG+fvkNi9flvHnENs5szODQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:51:24.836069Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.02122","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:cf0faa4f4abfe450f5d78fe9a3db1ee8b50f78811ce64973f75383f3c6f2dad7","sha256:d49696f263d5a4aefff30459fcab47e8b67da4b85c3cf02308fc8183642c9683"],"state_sha256":"0adbe03ffba2d9b1f4ef425dc1d470b5f7521e52980468f43268946a5c459e4d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WKuts/Jw05lwsQIEwqZV1h3pGybbbMuTObErT9NkWd3uf1y6CgdpumsEFm+CJF6KV1ipz3SM3YxpDAsnxgcKCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T01:53:07.501894Z","bundle_sha256":"e247a5955a51fc13e3d4bcff6928a1103df9327e8c62e83917c638200fb58376"}}