{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:U7CZZPD4UBMY2TKFMEKNWJQMUM","short_pith_number":"pith:U7CZZPD4","canonical_record":{"source":{"id":"2602.13940","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-15T00:31:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"89bdc1f13dc68f3fef867183c1971b6c602338a92211118b2e58f1ef8285a164","abstract_canon_sha256":"935047dee1d0007a5949d6d5c37d47953174c9b5570b249e02d103c6ba46e90c"},"schema_version":"1.0"},"canonical_sha256":"a7c59cbc7ca0598d4d456114db260ca325ff2c118b245e069e2a973fdfc331aa","source":{"kind":"arxiv","id":"2602.13940","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.13940","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"arxiv_version","alias_value":"2602.13940v2","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.13940","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"pith_short_12","alias_value":"U7CZZPD4UBMY","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"pith_short_16","alias_value":"U7CZZPD4UBMY2TKF","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"pith_short_8","alias_value":"U7CZZPD4","created_at":"2026-06-02T02:04:14Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:U7CZZPD4UBMY2TKFMEKNWJQMUM","target":"record","payload":{"canonical_record":{"source":{"id":"2602.13940","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-15T00:31:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"89bdc1f13dc68f3fef867183c1971b6c602338a92211118b2e58f1ef8285a164","abstract_canon_sha256":"935047dee1d0007a5949d6d5c37d47953174c9b5570b249e02d103c6ba46e90c"},"schema_version":"1.0"},"canonical_sha256":"a7c59cbc7ca0598d4d456114db260ca325ff2c118b245e069e2a973fdfc331aa","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:14.690603Z","signature_b64":"ANgEHiWAktl3gzsjVyeZCEZEk49007pXB9kiLkq3tYfQ5MmiXW9oqv3GN1heOnlyqlSaXRa4wYKXT0acX5QABA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a7c59cbc7ca0598d4d456114db260ca325ff2c118b245e069e2a973fdfc331aa","last_reissued_at":"2026-06-02T02:04:14.690183Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:14.690183Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.13940","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":"2026-06-02T02:04:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E7F31Jv+ZPtNCo2xleyy+Bc8/HTXgbdY8plHZFRY4iAIxBRo3lurKNXYcfhAdGOwI2lYFnZGVPM5la8HTmsoDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T13:40:47.735854Z"},"content_sha256":"c3f000f15b445baeffcf213c321cdfe9fd79f46110eed4bdfca5944f873cd03f","schema_version":"1.0","event_id":"sha256:c3f000f15b445baeffcf213c321cdfe9fd79f46110eed4bdfca5944f873cd03f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:U7CZZPD4UBMY2TKFMEKNWJQMUM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"You Can Learn Tokenization End-to-End with Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Roger Wattenhofer, Sam Dauncey","submitted_at":"2026-02-15T00:31:24Z","abstract_excerpt":"Tokenization is a hardcoded compression step which remains in the training pipeline of Large Language Models (LLMs), despite a general trend towards architectures becoming increasingly end-to-end. Prior work has shown promising results at scale in bringing this compression step inside the LLMs' architecture with heuristics to draw token boundaries, and also attempts to learn these token boundaries with straight-through estimates, which treat the problem of drawing discrete token boundaries as a continuous one. We show that these token boundaries can instead be learned using score function esti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.13940","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/2602.13940/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":"2026-06-02T02:04:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ye+MB62eOgGYN8s/rHAyywO5CPhoJjiearwx0lkEmWLtSGvRKtLVH700TedeBdNZTGEg7wwDAiyGSuuq4zGfDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T13:40:47.736461Z"},"content_sha256":"430c6ff79576dea921df5add4f8114464746b42d7b3df699b4b8ae3f277154c5","schema_version":"1.0","event_id":"sha256:430c6ff79576dea921df5add4f8114464746b42d7b3df699b4b8ae3f277154c5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/U7CZZPD4UBMY2TKFMEKNWJQMUM/bundle.json","state_url":"https://pith.science/pith/U7CZZPD4UBMY2TKFMEKNWJQMUM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/U7CZZPD4UBMY2TKFMEKNWJQMUM/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":"2026-08-04T13:40:47Z","links":{"resolver":"https://pith.science/pith/U7CZZPD4UBMY2TKFMEKNWJQMUM","bundle":"https://pith.science/pith/U7CZZPD4UBMY2TKFMEKNWJQMUM/bundle.json","state":"https://pith.science/pith/U7CZZPD4UBMY2TKFMEKNWJQMUM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/U7CZZPD4UBMY2TKFMEKNWJQMUM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:U7CZZPD4UBMY2TKFMEKNWJQMUM","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":"935047dee1d0007a5949d6d5c37d47953174c9b5570b249e02d103c6ba46e90c","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-15T00:31:24Z","title_canon_sha256":"89bdc1f13dc68f3fef867183c1971b6c602338a92211118b2e58f1ef8285a164"},"schema_version":"1.0","source":{"id":"2602.13940","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.13940","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"arxiv_version","alias_value":"2602.13940v2","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.13940","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"pith_short_12","alias_value":"U7CZZPD4UBMY","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"pith_short_16","alias_value":"U7CZZPD4UBMY2TKF","created_at":"2026-06-02T02:04:14Z"},{"alias_kind":"pith_short_8","alias_value":"U7CZZPD4","created_at":"2026-06-02T02:04:14Z"}],"graph_snapshots":[{"event_id":"sha256:430c6ff79576dea921df5add4f8114464746b42d7b3df699b4b8ae3f277154c5","target":"graph","created_at":"2026-06-02T02:04:14Z","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/2602.13940/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Tokenization is a hardcoded compression step which remains in the training pipeline of Large Language Models (LLMs), despite a general trend towards architectures becoming increasingly end-to-end. Prior work has shown promising results at scale in bringing this compression step inside the LLMs' architecture with heuristics to draw token boundaries, and also attempts to learn these token boundaries with straight-through estimates, which treat the problem of drawing discrete token boundaries as a continuous one. We show that these token boundaries can instead be learned using score function esti","authors_text":"Roger Wattenhofer, Sam Dauncey","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-15T00:31:24Z","title":"You Can Learn Tokenization End-to-End with Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.13940","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:c3f000f15b445baeffcf213c321cdfe9fd79f46110eed4bdfca5944f873cd03f","target":"record","created_at":"2026-06-02T02:04:14Z","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":"935047dee1d0007a5949d6d5c37d47953174c9b5570b249e02d103c6ba46e90c","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-15T00:31:24Z","title_canon_sha256":"89bdc1f13dc68f3fef867183c1971b6c602338a92211118b2e58f1ef8285a164"},"schema_version":"1.0","source":{"id":"2602.13940","kind":"arxiv","version":2}},"canonical_sha256":"a7c59cbc7ca0598d4d456114db260ca325ff2c118b245e069e2a973fdfc331aa","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a7c59cbc7ca0598d4d456114db260ca325ff2c118b245e069e2a973fdfc331aa","first_computed_at":"2026-06-02T02:04:14.690183Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T02:04:14.690183Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ANgEHiWAktl3gzsjVyeZCEZEk49007pXB9kiLkq3tYfQ5MmiXW9oqv3GN1heOnlyqlSaXRa4wYKXT0acX5QABA==","signature_status":"signed_v1","signed_at":"2026-06-02T02:04:14.690603Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.13940","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c3f000f15b445baeffcf213c321cdfe9fd79f46110eed4bdfca5944f873cd03f","sha256:430c6ff79576dea921df5add4f8114464746b42d7b3df699b4b8ae3f277154c5"],"state_sha256":"a0ff722e35711204c8f0bd16d9d5101e419912f282841e9482fa95178789fe15"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VB4aM7gnQ/wkBTj3A2HZFRPRQQ9HpumrjGD7y7wUUr2QhsNq83gzZJ7kJFQjHnZR9U3iPR8qzrmnchBbSxrWBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T13:40:47.740509Z","bundle_sha256":"42b3e15079361dc5aefe03a2999d65b9548afd3643bd1da864413aff6fb4311d"}}