{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3NWEWSSICNMPHJEDSNLUH3EBWG","short_pith_number":"pith:3NWEWSSI","canonical_record":{"source":{"id":"2505.15548","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-21T14:12:30Z","cross_cats_sorted":[],"title_canon_sha256":"52ea267f3832becc485518bdc00a27f25509be2565db08dda11e9145d691b433","abstract_canon_sha256":"6da6629ee24ef03aa5bea54503761d7f586cc3a6159e223ca673126a13b150e4"},"schema_version":"1.0"},"canonical_sha256":"db6c4b4a481358f3a483935743ec81b1b4a3b8efef7f7ecbb45de09ff550e5bb","source":{"kind":"arxiv","id":"2505.15548","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.15548","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"arxiv_version","alias_value":"2505.15548v1","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.15548","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"pith_short_12","alias_value":"3NWEWSSICNMP","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"pith_short_16","alias_value":"3NWEWSSICNMPHJED","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"pith_short_8","alias_value":"3NWEWSSI","created_at":"2026-07-05T11:06:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3NWEWSSICNMPHJEDSNLUH3EBWG","target":"record","payload":{"canonical_record":{"source":{"id":"2505.15548","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-21T14:12:30Z","cross_cats_sorted":[],"title_canon_sha256":"52ea267f3832becc485518bdc00a27f25509be2565db08dda11e9145d691b433","abstract_canon_sha256":"6da6629ee24ef03aa5bea54503761d7f586cc3a6159e223ca673126a13b150e4"},"schema_version":"1.0"},"canonical_sha256":"db6c4b4a481358f3a483935743ec81b1b4a3b8efef7f7ecbb45de09ff550e5bb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:47.554760Z","signature_b64":"MpJyeos4BS8ZOlEC6eVyLy3kF45abtbbGnHZI8mf0Nnjq2MZljJ6ASfZ/vDMsDOjRFFmb90QhSSr2g1gHBiKDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db6c4b4a481358f3a483935743ec81b1b4a3b8efef7f7ecbb45de09ff550e5bb","last_reissued_at":"2026-07-05T11:06:47.554268Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:47.554268Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.15548","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":"2026-07-05T11:06:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YO6oSxfWo6K/eOD19I9xrXuu0x66M4JdsZ5zy0rwaKXPKoK/gQMyysTuPQqjP5UWoHgC5fI5CYAUr1UNYJ1kDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T20:34:13.181958Z"},"content_sha256":"ba57306ead29c5ceccbf1622410d34878099e102e560bebe1eca5387641fac7e","schema_version":"1.0","event_id":"sha256:ba57306ead29c5ceccbf1622410d34878099e102e560bebe1eca5387641fac7e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3NWEWSSICNMPHJEDSNLUH3EBWG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Short-Range Dependency Effects on Transformer Instability and a Decomposed Attention Solution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Suvadeep Hajra","submitted_at":"2025-05-21T14:12:30Z","abstract_excerpt":"Transformer language models have driven significant progress across various fields, including natural language processing and computer vision. A central component of these models is the self-attention (SA) mechanism, which learns rich vector representations of tokens by modeling their relationships with others in a sequence. However, despite extensive research, transformers continue to suffer from training instability -- often manifesting as spikes or divergence in the training loss during a run.\n  In this work, we identify one source of this instability: SA's limited ability to capture short-"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.15548","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/2505.15548/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":"2026-07-05T11:06:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YYtVgXwHBv8H3o9sKRhBvE0rkeD00Y0KPRKMjPF0YlLhFS+Yugjnt2M1eLy8PhvtHvv2a+wk03B0xApOZD9oCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T20:34:13.182522Z"},"content_sha256":"5b734ca537bbabfa3243541b913a4e3c58974381c6c5d30b43f9f16305ebf320","schema_version":"1.0","event_id":"sha256:5b734ca537bbabfa3243541b913a4e3c58974381c6c5d30b43f9f16305ebf320"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3NWEWSSICNMPHJEDSNLUH3EBWG/bundle.json","state_url":"https://pith.science/pith/3NWEWSSICNMPHJEDSNLUH3EBWG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3NWEWSSICNMPHJEDSNLUH3EBWG/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":"2026-08-09T20:34:13Z","links":{"resolver":"https://pith.science/pith/3NWEWSSICNMPHJEDSNLUH3EBWG","bundle":"https://pith.science/pith/3NWEWSSICNMPHJEDSNLUH3EBWG/bundle.json","state":"https://pith.science/pith/3NWEWSSICNMPHJEDSNLUH3EBWG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3NWEWSSICNMPHJEDSNLUH3EBWG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3NWEWSSICNMPHJEDSNLUH3EBWG","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":"6da6629ee24ef03aa5bea54503761d7f586cc3a6159e223ca673126a13b150e4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-21T14:12:30Z","title_canon_sha256":"52ea267f3832becc485518bdc00a27f25509be2565db08dda11e9145d691b433"},"schema_version":"1.0","source":{"id":"2505.15548","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.15548","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"arxiv_version","alias_value":"2505.15548v1","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.15548","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"pith_short_12","alias_value":"3NWEWSSICNMP","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"pith_short_16","alias_value":"3NWEWSSICNMPHJED","created_at":"2026-07-05T11:06:47Z"},{"alias_kind":"pith_short_8","alias_value":"3NWEWSSI","created_at":"2026-07-05T11:06:47Z"}],"graph_snapshots":[{"event_id":"sha256:5b734ca537bbabfa3243541b913a4e3c58974381c6c5d30b43f9f16305ebf320","target":"graph","created_at":"2026-07-05T11:06:47Z","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/2505.15548/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Transformer language models have driven significant progress across various fields, including natural language processing and computer vision. A central component of these models is the self-attention (SA) mechanism, which learns rich vector representations of tokens by modeling their relationships with others in a sequence. However, despite extensive research, transformers continue to suffer from training instability -- often manifesting as spikes or divergence in the training loss during a run.\n  In this work, we identify one source of this instability: SA's limited ability to capture short-","authors_text":"Suvadeep Hajra","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-21T14:12:30Z","title":"Short-Range Dependency Effects on Transformer Instability and a Decomposed Attention Solution"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.15548","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:ba57306ead29c5ceccbf1622410d34878099e102e560bebe1eca5387641fac7e","target":"record","created_at":"2026-07-05T11:06:47Z","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":"6da6629ee24ef03aa5bea54503761d7f586cc3a6159e223ca673126a13b150e4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-21T14:12:30Z","title_canon_sha256":"52ea267f3832becc485518bdc00a27f25509be2565db08dda11e9145d691b433"},"schema_version":"1.0","source":{"id":"2505.15548","kind":"arxiv","version":1}},"canonical_sha256":"db6c4b4a481358f3a483935743ec81b1b4a3b8efef7f7ecbb45de09ff550e5bb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"db6c4b4a481358f3a483935743ec81b1b4a3b8efef7f7ecbb45de09ff550e5bb","first_computed_at":"2026-07-05T11:06:47.554268Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:06:47.554268Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"MpJyeos4BS8ZOlEC6eVyLy3kF45abtbbGnHZI8mf0Nnjq2MZljJ6ASfZ/vDMsDOjRFFmb90QhSSr2g1gHBiKDA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:06:47.554760Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.15548","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ba57306ead29c5ceccbf1622410d34878099e102e560bebe1eca5387641fac7e","sha256:5b734ca537bbabfa3243541b913a4e3c58974381c6c5d30b43f9f16305ebf320"],"state_sha256":"d0b9f48995c1fc65d2199a898f9000ac2167826d2963453684938bf269a3b9f3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8gDuZhuQBO9m9eXp5MfVPdcOqLhq501VBbcvaiZdo9iTw0my0k3I1EX4t9IHk0qH6XAVoewz7sY/6NSMvRvoCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T20:34:13.188025Z","bundle_sha256":"508babcfd6de89045463eb032d452e934455987cfa67a31db01ae461b532e3fc"}}