{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:LIJM6NQO3E7AP3CMCPQNIP2VK2","short_pith_number":"pith:LIJM6NQO","canonical_record":{"source":{"id":"2503.09202","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-12T09:46:59Z","cross_cats_sorted":[],"title_canon_sha256":"41decc99981a8c43e53a04dd4cead0ba6a08a5085227ed04a00a8afeb8f922d4","abstract_canon_sha256":"17c47bc9f36e42d6a4f2d2e06bf462276e3e23cd1056c6b4088d4c32da5b1137"},"schema_version":"1.0"},"canonical_sha256":"5a12cf360ed93e07ec4c13e0d43f55568655c07fab0770e3a71944aebca7debd","source":{"kind":"arxiv","id":"2503.09202","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.09202","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"arxiv_version","alias_value":"2503.09202v1","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.09202","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"pith_short_12","alias_value":"LIJM6NQO3E7A","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"pith_short_16","alias_value":"LIJM6NQO3E7AP3CM","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"pith_short_8","alias_value":"LIJM6NQO","created_at":"2026-07-05T10:29:51Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:LIJM6NQO3E7AP3CMCPQNIP2VK2","target":"record","payload":{"canonical_record":{"source":{"id":"2503.09202","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-12T09:46:59Z","cross_cats_sorted":[],"title_canon_sha256":"41decc99981a8c43e53a04dd4cead0ba6a08a5085227ed04a00a8afeb8f922d4","abstract_canon_sha256":"17c47bc9f36e42d6a4f2d2e06bf462276e3e23cd1056c6b4088d4c32da5b1137"},"schema_version":"1.0"},"canonical_sha256":"5a12cf360ed93e07ec4c13e0d43f55568655c07fab0770e3a71944aebca7debd","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:29:51.378898Z","signature_b64":"ZmQO3o3hRTLQeeK73POdFWfDCuOApPW2B5OLLMpXubbbHz08csgOv/SW2xWBmJsEIgArFeg3dxWNV/X+C+ZwBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a12cf360ed93e07ec4c13e0d43f55568655c07fab0770e3a71944aebca7debd","last_reissued_at":"2026-07-05T10:29:51.378059Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:29:51.378059Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.09202","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":"2026-07-05T10:29:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"y0XQZkA66pN4lA278NonsVjcyunsbbW135irKoIJz8t+M/T/AynncEMcyMKp6yfB8TtpcqxKhzKQ68g0P4FyCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T14:12:12.833758Z"},"content_sha256":"a2880ce009ea54973b0c39865bd341c215606c0913747b0ca49b0709aa7df9a2","schema_version":"1.0","event_id":"sha256:a2880ce009ea54973b0c39865bd341c215606c0913747b0ca49b0709aa7df9a2"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:LIJM6NQO3E7AP3CMCPQNIP2VK2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Token Weighting for Long-Range Language Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Falko Helm, Iryna Gurevych, Nico Daheim","submitted_at":"2025-03-12T09:46:59Z","abstract_excerpt":"Many applications of large language models (LLMs) require long-context understanding, but models continue to struggle with such tasks. We hypothesize that conventional next-token prediction training could contribute to this, because each token is assigned equal weight. Yet, intuitively, the amount of context needed to predict the next token accurately varies greatly across different data. To reflect this, we propose various novel token-weighting schemes that assign different weights to each training token in the loss, thereby generalizing existing works. For this, we categorize token-weighting"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.09202","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/2503.09202/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":"2026-07-05T10:29:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"F/vqyQdACOJvr5WTJse6xM0Iiq+62EIr0AcUhng+EIQqtdfOu0m52MGQc2qlUlEt97d11K0PhnGHkxnfT5WDCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T14:12:12.834366Z"},"content_sha256":"a43f60eb732304a5b4e32cb16ae094b934951d846e3f615bc9a059e548386b4b","schema_version":"1.0","event_id":"sha256:a43f60eb732304a5b4e32cb16ae094b934951d846e3f615bc9a059e548386b4b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LIJM6NQO3E7AP3CMCPQNIP2VK2/bundle.json","state_url":"https://pith.science/pith/LIJM6NQO3E7AP3CMCPQNIP2VK2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LIJM6NQO3E7AP3CMCPQNIP2VK2/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":"2026-08-07T14:12:12Z","links":{"resolver":"https://pith.science/pith/LIJM6NQO3E7AP3CMCPQNIP2VK2","bundle":"https://pith.science/pith/LIJM6NQO3E7AP3CMCPQNIP2VK2/bundle.json","state":"https://pith.science/pith/LIJM6NQO3E7AP3CMCPQNIP2VK2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LIJM6NQO3E7AP3CMCPQNIP2VK2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:LIJM6NQO3E7AP3CMCPQNIP2VK2","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":"17c47bc9f36e42d6a4f2d2e06bf462276e3e23cd1056c6b4088d4c32da5b1137","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-12T09:46:59Z","title_canon_sha256":"41decc99981a8c43e53a04dd4cead0ba6a08a5085227ed04a00a8afeb8f922d4"},"schema_version":"1.0","source":{"id":"2503.09202","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.09202","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"arxiv_version","alias_value":"2503.09202v1","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.09202","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"pith_short_12","alias_value":"LIJM6NQO3E7A","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"pith_short_16","alias_value":"LIJM6NQO3E7AP3CM","created_at":"2026-07-05T10:29:51Z"},{"alias_kind":"pith_short_8","alias_value":"LIJM6NQO","created_at":"2026-07-05T10:29:51Z"}],"graph_snapshots":[{"event_id":"sha256:a43f60eb732304a5b4e32cb16ae094b934951d846e3f615bc9a059e548386b4b","target":"graph","created_at":"2026-07-05T10:29:51Z","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/2503.09202/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Many applications of large language models (LLMs) require long-context understanding, but models continue to struggle with such tasks. We hypothesize that conventional next-token prediction training could contribute to this, because each token is assigned equal weight. Yet, intuitively, the amount of context needed to predict the next token accurately varies greatly across different data. To reflect this, we propose various novel token-weighting schemes that assign different weights to each training token in the loss, thereby generalizing existing works. For this, we categorize token-weighting","authors_text":"Falko Helm, Iryna Gurevych, Nico Daheim","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-12T09:46:59Z","title":"Token Weighting for Long-Range Language Modeling"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.09202","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:a2880ce009ea54973b0c39865bd341c215606c0913747b0ca49b0709aa7df9a2","target":"record","created_at":"2026-07-05T10:29:51Z","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":"17c47bc9f36e42d6a4f2d2e06bf462276e3e23cd1056c6b4088d4c32da5b1137","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-12T09:46:59Z","title_canon_sha256":"41decc99981a8c43e53a04dd4cead0ba6a08a5085227ed04a00a8afeb8f922d4"},"schema_version":"1.0","source":{"id":"2503.09202","kind":"arxiv","version":1}},"canonical_sha256":"5a12cf360ed93e07ec4c13e0d43f55568655c07fab0770e3a71944aebca7debd","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5a12cf360ed93e07ec4c13e0d43f55568655c07fab0770e3a71944aebca7debd","first_computed_at":"2026-07-05T10:29:51.378059Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:29:51.378059Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZmQO3o3hRTLQeeK73POdFWfDCuOApPW2B5OLLMpXubbbHz08csgOv/SW2xWBmJsEIgArFeg3dxWNV/X+C+ZwBg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:29:51.378898Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.09202","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a2880ce009ea54973b0c39865bd341c215606c0913747b0ca49b0709aa7df9a2","sha256:a43f60eb732304a5b4e32cb16ae094b934951d846e3f615bc9a059e548386b4b"],"state_sha256":"54ac0ab434bec56009e7602690cbfc7336584171eea5ded7d72b25f3c9609e94"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"x4TBSIcwV/o7nBn5PK0r58L3gfJS/tcTSX0aajQ47mpLSpFB+Y6oZal6IHfAA33Epe4Xrv55sV2aDNsmMjJCCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T14:12:12.839291Z","bundle_sha256":"d311e496d7ee660f6f31e317eb803b794bb4c67a225d7d7bc025cdb89d163546"}}