{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:X5R6XHVJKOLO56LFPET4YJ5P5A","short_pith_number":"pith:X5R6XHVJ","canonical_record":{"source":{"id":"2605.14589","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-14T09:00:03Z","cross_cats_sorted":[],"title_canon_sha256":"ed77087556ed0d48317c2d5798de214de2707e733277664f483df5c8217b5ab2","abstract_canon_sha256":"49a76783063716cabb3487ccd241489241ded9027beb749d517950142bd96167"},"schema_version":"1.0"},"canonical_sha256":"bf63eb9ea95396eef9657927cc27afe8040909cfdacdb83c1c8df3399b9b9c5b","source":{"kind":"arxiv","id":"2605.14589","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14589","created_at":"2026-05-17T23:39:05Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14589v1","created_at":"2026-05-17T23:39:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14589","created_at":"2026-05-17T23:39:05Z"},{"alias_kind":"pith_short_12","alias_value":"X5R6XHVJKOLO","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"X5R6XHVJKOLO56LF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"X5R6XHVJ","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:X5R6XHVJKOLO56LFPET4YJ5P5A","target":"record","payload":{"canonical_record":{"source":{"id":"2605.14589","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-14T09:00:03Z","cross_cats_sorted":[],"title_canon_sha256":"ed77087556ed0d48317c2d5798de214de2707e733277664f483df5c8217b5ab2","abstract_canon_sha256":"49a76783063716cabb3487ccd241489241ded9027beb749d517950142bd96167"},"schema_version":"1.0"},"canonical_sha256":"bf63eb9ea95396eef9657927cc27afe8040909cfdacdb83c1c8df3399b9b9c5b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:05.268746Z","signature_b64":"gYx0hi/QcU9BzLVqpHGVzR8Piw+oLu/SBI6+4rAQO/rtlNk5joFYM4guLkzA85pSTo7o4SlnkM1P4iRTZ5zhDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bf63eb9ea95396eef9657927cc27afe8040909cfdacdb83c1c8df3399b9b9c5b","last_reissued_at":"2026-05-17T23:39:05.268119Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:05.268119Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.14589","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":"2026-05-17T23:39:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SkpNY2yxRXWpdlx5NG4u3cE6sW1qn4Dnkk7FOqUg/5D3ifVu3JAPO/IK9c9ai4ZlH0nW1Un/yFiw6PKH00+vAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T12:14:37.490490Z"},"content_sha256":"bf51f0077e1a2bf3b573aab146e4e37aeb7df7bf6c281a75e46f3c26ff914189","schema_version":"1.0","event_id":"sha256:bf51f0077e1a2bf3b573aab146e4e37aeb7df7bf6c281a75e46f3c26ff914189"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:X5R6XHVJKOLO56LFPET4YJ5P5A","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"EndPrompt: Efficient Long-Context Extension via Terminal Anchoring","license":"http://creativecommons.org/licenses/by/4.0/","headline":"EndPrompt extends LLM context windows to 64K by training only on short sequences with a terminal prompt anchored at target positions.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dawei Yin, Fang Wang, Han Tian, Haoyi Xiong, Jiamin Chen, Jiashu Zhao, Jinman Zhao, Luxuan Chen, Rui Kong, Shuaiqiang Wang, Xinran Chen, Yuchen Li","submitted_at":"2026-05-14T09:00:03Z","abstract_excerpt":"Extending the context window of large language models typically requires training on sequences at the target length, incurring quadratic memory and computational costs that make long-context adaptation expensive and difficult to reproduce. We propose EndPrompt, a method that achieves effective context extension using only short training sequences. The core insight is that exposing a model to long-range relative positional distances does not require constructing full-length inputs: we preserve the original short context as an intact first segment and append a brief terminal prompt as a second s"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"long-context generalization can be induced from sparse positional supervision, challenging the prevailing assumption that dense long-sequence training is necessary for reliable context-window extension.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That assigning positional indices near the target length to a brief terminal prompt in short sequences preserves the necessary relative distances and semantic continuity for effective long-context learning without introducing artifacts from the artificial split.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"EndPrompt induces reliable long-context generalization in LLaMA models from sparse positional supervision via a two-segment short-sequence construction with terminal anchoring.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"EndPrompt extends LLM context windows to 64K by training only on short sequences with a terminal prompt anchored at target positions.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"956f40ed766a3a7c86082e6634ac42ae553e431830d2871c7e3df584ff326ee9"},"source":{"id":"2605.14589","kind":"arxiv","version":1},"verdict":{"id":"9d7ff822-5310-44e5-80d4-89377fc88121","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T01:26:59.577618Z","strongest_claim":"long-context generalization can be induced from sparse positional supervision, challenging the prevailing assumption that dense long-sequence training is necessary for reliable context-window extension.","one_line_summary":"EndPrompt induces reliable long-context generalization in LLaMA models from sparse positional supervision via a two-segment short-sequence construction with terminal anchoring.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That assigning positional indices near the target length to a brief terminal prompt in short sequences preserves the necessary relative distances and semantic continuity for effective long-context learning without introducing artifacts from the artificial split.","pith_extraction_headline":"EndPrompt extends LLM context windows to 64K by training only on short sequences with a terminal prompt anchored at target positions."},"references":{"count":35,"sample":[{"doi":"","year":2024,"title":"Longalign: A recipe for long context alignment of large language models","work_id":"aa055c89-c8e0-44aa-ad75-65f83f864a16","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding","work_id":"ba7831c4-9427-4e0e-a5c1-4e98511f4b53","ref_index":2,"cited_arxiv_id":"2308.14508","is_internal_anchor":true},{"doi":"","year":2004,"title":"Longformer: The Long-Document Transformer","work_id":"abea7a44-6668-4de7-aab6-f53a6e5aa088","ref_index":3,"cited_arxiv_id":"2004.05150","is_internal_anchor":true},{"doi":"","year":2022,"title":"Lexglue: A benchmark dataset for legal language understanding in english.Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics, 2022","work_id":"012aa679-98a0-4ee5-84f8-069cf54e0fb1","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"L-Eval: Instituting standardized evaluation for long context language models","work_id":"8c5b99fd-55bf-4d88-b148-8bb61e28ee99","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":35,"snapshot_sha256":"ce7bbf1d451921abfe095d0466d9340d67f6b3022fa57d4a5f8864c0facbd299","internal_anchors":10},"formal_canon":{"evidence_count":2,"snapshot_sha256":"60d9ddabfd61ffc3d0686fc1dbee92877d5e22a38ebf902babd40f08109f16c3"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"9d7ff822-5310-44e5-80d4-89377fc88121"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E1Paho8Mgz4xvRarmi2CH6VSJLgYEcMNziQ0/CtKIXWGaZOvn/ARULxoKycqN+LOp+/Lt8xzF/nrG6hqqcDGDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T12:14:37.491509Z"},"content_sha256":"f853049c30aa08f12b9fbef06191efb9f2b548fa8724a69d86edeb0908b93a88","schema_version":"1.0","event_id":"sha256:f853049c30aa08f12b9fbef06191efb9f2b548fa8724a69d86edeb0908b93a88"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/X5R6XHVJKOLO56LFPET4YJ5P5A/bundle.json","state_url":"https://pith.science/pith/X5R6XHVJKOLO56LFPET4YJ5P5A/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/X5R6XHVJKOLO56LFPET4YJ5P5A/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":"2026-05-30T12:14:37Z","links":{"resolver":"https://pith.science/pith/X5R6XHVJKOLO56LFPET4YJ5P5A","bundle":"https://pith.science/pith/X5R6XHVJKOLO56LFPET4YJ5P5A/bundle.json","state":"https://pith.science/pith/X5R6XHVJKOLO56LFPET4YJ5P5A/state.json","well_known_bundle":"https://pith.science/.well-known/pith/X5R6XHVJKOLO56LFPET4YJ5P5A/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:X5R6XHVJKOLO56LFPET4YJ5P5A","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":"49a76783063716cabb3487ccd241489241ded9027beb749d517950142bd96167","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-14T09:00:03Z","title_canon_sha256":"ed77087556ed0d48317c2d5798de214de2707e733277664f483df5c8217b5ab2"},"schema_version":"1.0","source":{"id":"2605.14589","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14589","created_at":"2026-05-17T23:39:05Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14589v1","created_at":"2026-05-17T23:39:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14589","created_at":"2026-05-17T23:39:05Z"},{"alias_kind":"pith_short_12","alias_value":"X5R6XHVJKOLO","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"X5R6XHVJKOLO56LF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"X5R6XHVJ","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:f853049c30aa08f12b9fbef06191efb9f2b548fa8724a69d86edeb0908b93a88","target":"graph","created_at":"2026-05-17T23:39:05Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"long-context generalization can be induced from sparse positional supervision, challenging the prevailing assumption that dense long-sequence training is necessary for reliable context-window extension."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That assigning positional indices near the target length to a brief terminal prompt in short sequences preserves the necessary relative distances and semantic continuity for effective long-context learning without introducing artifacts from the artificial split."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"EndPrompt induces reliable long-context generalization in LLaMA models from sparse positional supervision via a two-segment short-sequence construction with terminal anchoring."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"EndPrompt extends LLM context windows to 64K by training only on short sequences with a terminal prompt anchored at target positions."}],"snapshot_sha256":"956f40ed766a3a7c86082e6634ac42ae553e431830d2871c7e3df584ff326ee9"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"60d9ddabfd61ffc3d0686fc1dbee92877d5e22a38ebf902babd40f08109f16c3"},"paper":{"abstract_excerpt":"Extending the context window of large language models typically requires training on sequences at the target length, incurring quadratic memory and computational costs that make long-context adaptation expensive and difficult to reproduce. We propose EndPrompt, a method that achieves effective context extension using only short training sequences. The core insight is that exposing a model to long-range relative positional distances does not require constructing full-length inputs: we preserve the original short context as an intact first segment and append a brief terminal prompt as a second s","authors_text":"Dawei Yin, Fang Wang, Han Tian, Haoyi Xiong, Jiamin Chen, Jiashu Zhao, Jinman Zhao, Luxuan Chen, Rui Kong, Shuaiqiang Wang, Xinran Chen, Yuchen Li","cross_cats":[],"headline":"EndPrompt extends LLM context windows to 64K by training only on short sequences with a terminal prompt anchored at target positions.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-14T09:00:03Z","title":"EndPrompt: Efficient Long-Context Extension via Terminal Anchoring"},"references":{"count":35,"internal_anchors":10,"resolved_work":35,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Longalign: A recipe for long context alignment of large language models","work_id":"aa055c89-c8e0-44aa-ad75-65f83f864a16","year":2024},{"cited_arxiv_id":"2308.14508","doi":"","is_internal_anchor":true,"ref_index":2,"title":"LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding","work_id":"ba7831c4-9427-4e0e-a5c1-4e98511f4b53","year":2023},{"cited_arxiv_id":"2004.05150","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Longformer: The Long-Document Transformer","work_id":"abea7a44-6668-4de7-aab6-f53a6e5aa088","year":2004},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Lexglue: A benchmark dataset for legal language understanding in english.Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics, 2022","work_id":"012aa679-98a0-4ee5-84f8-069cf54e0fb1","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"L-Eval: Instituting standardized evaluation for long context language models","work_id":"8c5b99fd-55bf-4d88-b148-8bb61e28ee99","year":2023}],"snapshot_sha256":"ce7bbf1d451921abfe095d0466d9340d67f6b3022fa57d4a5f8864c0facbd299"},"source":{"id":"2605.14589","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T01:26:59.577618Z","id":"9d7ff822-5310-44e5-80d4-89377fc88121","model_set":{"reader":"grok-4.3"},"one_line_summary":"EndPrompt induces reliable long-context generalization in LLaMA models from sparse positional supervision via a two-segment short-sequence construction with terminal anchoring.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"EndPrompt extends LLM context windows to 64K by training only on short sequences with a terminal prompt anchored at target positions.","strongest_claim":"long-context generalization can be induced from sparse positional supervision, challenging the prevailing assumption that dense long-sequence training is necessary for reliable context-window extension.","weakest_assumption":"That assigning positional indices near the target length to a brief terminal prompt in short sequences preserves the necessary relative distances and semantic continuity for effective long-context learning without introducing artifacts from the artificial split."}},"verdict_id":"9d7ff822-5310-44e5-80d4-89377fc88121"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:bf51f0077e1a2bf3b573aab146e4e37aeb7df7bf6c281a75e46f3c26ff914189","target":"record","created_at":"2026-05-17T23:39:05Z","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":"49a76783063716cabb3487ccd241489241ded9027beb749d517950142bd96167","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-14T09:00:03Z","title_canon_sha256":"ed77087556ed0d48317c2d5798de214de2707e733277664f483df5c8217b5ab2"},"schema_version":"1.0","source":{"id":"2605.14589","kind":"arxiv","version":1}},"canonical_sha256":"bf63eb9ea95396eef9657927cc27afe8040909cfdacdb83c1c8df3399b9b9c5b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bf63eb9ea95396eef9657927cc27afe8040909cfdacdb83c1c8df3399b9b9c5b","first_computed_at":"2026-05-17T23:39:05.268119Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:05.268119Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gYx0hi/QcU9BzLVqpHGVzR8Piw+oLu/SBI6+4rAQO/rtlNk5joFYM4guLkzA85pSTo7o4SlnkM1P4iRTZ5zhDA==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:05.268746Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.14589","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bf51f0077e1a2bf3b573aab146e4e37aeb7df7bf6c281a75e46f3c26ff914189","sha256:f853049c30aa08f12b9fbef06191efb9f2b548fa8724a69d86edeb0908b93a88"],"state_sha256":"c2028370e62512c9fad38ab0b2565ddf95084832a19b8c0909cfeaed96f6fdec"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"o+XcWoj0VrvHZ0Zsx8PzgzLOinVjPELEZAfovdRT/xEFOhTr/91ehlTUBYsWrddJdIVq1deNeIJnM36DRhsNDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T12:14:37.496124Z","bundle_sha256":"4619778154f3134b442070322a4eea919a423533a158301d032c83f2e94441e2"}}