{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:XIUH7MZC4Q4WJMGTYYJ2AH4QRE","short_pith_number":"pith:XIUH7MZC","canonical_record":{"source":{"id":"2106.14881","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-06-28T17:59:33Z","cross_cats_sorted":[],"title_canon_sha256":"96dcd19035b316638979c7a18193c6ab4c5b0a9dbb5fa6b789f962ac2795b44a","abstract_canon_sha256":"19834498e30aafa6442bb3c3e5d69ef58b44b5383ad7cf3bc504f004e1718e40"},"schema_version":"1.0"},"canonical_sha256":"ba287fb322e43964b0d3c613a01f908926359a657be37a7c719619436edaf81a","source":{"kind":"arxiv","id":"2106.14881","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2106.14881","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"arxiv_version","alias_value":"2106.14881v3","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.14881","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"pith_short_12","alias_value":"XIUH7MZC4Q4W","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"pith_short_16","alias_value":"XIUH7MZC4Q4WJMGT","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"pith_short_8","alias_value":"XIUH7MZC","created_at":"2026-07-05T03:25:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:XIUH7MZC4Q4WJMGTYYJ2AH4QRE","target":"record","payload":{"canonical_record":{"source":{"id":"2106.14881","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-06-28T17:59:33Z","cross_cats_sorted":[],"title_canon_sha256":"96dcd19035b316638979c7a18193c6ab4c5b0a9dbb5fa6b789f962ac2795b44a","abstract_canon_sha256":"19834498e30aafa6442bb3c3e5d69ef58b44b5383ad7cf3bc504f004e1718e40"},"schema_version":"1.0"},"canonical_sha256":"ba287fb322e43964b0d3c613a01f908926359a657be37a7c719619436edaf81a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:25:32.497340Z","signature_b64":"01v487iPDF/r3iNBWuFK5O5uaScTEGzBPb/pTGLrLvEzLXCXjiK1hG0TJqlEUDUtIogfnny58PN8COMyjFbeBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba287fb322e43964b0d3c613a01f908926359a657be37a7c719619436edaf81a","last_reissued_at":"2026-07-05T03:25:32.496944Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:25:32.496944Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2106.14881","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T03:25:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ty667otcM/J6BaKGaUo4l8amAxlIubIZUsU6v8uoOcjIugnThcm0EA3cPS2vXb06p7Uu+Hq+jP6Bq9MssG8ZAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T00:35:31.923131Z"},"content_sha256":"a8b512a08b1b8b47e6b42fe4cd6da607b69ce8a9ea1ada35a22d7ddca2477eb0","schema_version":"1.0","event_id":"sha256:a8b512a08b1b8b47e6b42fe4cd6da607b69ce8a9ea1ada35a22d7ddca2477eb0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:XIUH7MZC4Q4WJMGTYYJ2AH4QRE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Early Convolutions Help Transformers See Better","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Eric Mintun, Mannat Singh, Piotr Doll\\'ar, Ross Girshick, Tete Xiao, Trevor Darrell","submitted_at":"2021-06-28T17:59:33Z","abstract_excerpt":"Vision transformer (ViT) models exhibit substandard optimizability. In particular, they are sensitive to the choice of optimizer (AdamW vs. SGD), optimizer hyperparameters, and training schedule length. In comparison, modern convolutional neural networks are easier to optimize. Why is this the case? In this work, we conjecture that the issue lies with the patchify stem of ViT models, which is implemented by a stride-p p*p convolution (p=16 by default) applied to the input image. This large-kernel plus large-stride convolution runs counter to typical design choices of convolutional layers in ne"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.14881","kind":"arxiv","version":3},"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/2106.14881/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":"2026-07-05T03:25:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+a6bIEhlYpPAbsW1fIJUCnsMY4uAEAhD3PQfx6dDImUOg48oxkBIf9urPK+hmHUoyzrImw1OqKmiOYC/WNngCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T00:35:31.923611Z"},"content_sha256":"7a13ba1a2da31c2e0e9c3a4cb1fcde89e17037ce0518d64a21d8b9a58f94dc3b","schema_version":"1.0","event_id":"sha256:7a13ba1a2da31c2e0e9c3a4cb1fcde89e17037ce0518d64a21d8b9a58f94dc3b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XIUH7MZC4Q4WJMGTYYJ2AH4QRE/bundle.json","state_url":"https://pith.science/pith/XIUH7MZC4Q4WJMGTYYJ2AH4QRE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XIUH7MZC4Q4WJMGTYYJ2AH4QRE/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":"2026-08-10T00:35:31Z","links":{"resolver":"https://pith.science/pith/XIUH7MZC4Q4WJMGTYYJ2AH4QRE","bundle":"https://pith.science/pith/XIUH7MZC4Q4WJMGTYYJ2AH4QRE/bundle.json","state":"https://pith.science/pith/XIUH7MZC4Q4WJMGTYYJ2AH4QRE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XIUH7MZC4Q4WJMGTYYJ2AH4QRE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:XIUH7MZC4Q4WJMGTYYJ2AH4QRE","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":"19834498e30aafa6442bb3c3e5d69ef58b44b5383ad7cf3bc504f004e1718e40","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-06-28T17:59:33Z","title_canon_sha256":"96dcd19035b316638979c7a18193c6ab4c5b0a9dbb5fa6b789f962ac2795b44a"},"schema_version":"1.0","source":{"id":"2106.14881","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2106.14881","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"arxiv_version","alias_value":"2106.14881v3","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.14881","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"pith_short_12","alias_value":"XIUH7MZC4Q4W","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"pith_short_16","alias_value":"XIUH7MZC4Q4WJMGT","created_at":"2026-07-05T03:25:32Z"},{"alias_kind":"pith_short_8","alias_value":"XIUH7MZC","created_at":"2026-07-05T03:25:32Z"}],"graph_snapshots":[{"event_id":"sha256:7a13ba1a2da31c2e0e9c3a4cb1fcde89e17037ce0518d64a21d8b9a58f94dc3b","target":"graph","created_at":"2026-07-05T03:25:32Z","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/2106.14881/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Vision transformer (ViT) models exhibit substandard optimizability. In particular, they are sensitive to the choice of optimizer (AdamW vs. SGD), optimizer hyperparameters, and training schedule length. In comparison, modern convolutional neural networks are easier to optimize. Why is this the case? In this work, we conjecture that the issue lies with the patchify stem of ViT models, which is implemented by a stride-p p*p convolution (p=16 by default) applied to the input image. This large-kernel plus large-stride convolution runs counter to typical design choices of convolutional layers in ne","authors_text":"Eric Mintun, Mannat Singh, Piotr Doll\\'ar, Ross Girshick, Tete Xiao, Trevor Darrell","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-06-28T17:59:33Z","title":"Early Convolutions Help Transformers See Better"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.14881","kind":"arxiv","version":3},"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:a8b512a08b1b8b47e6b42fe4cd6da607b69ce8a9ea1ada35a22d7ddca2477eb0","target":"record","created_at":"2026-07-05T03:25:32Z","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":"19834498e30aafa6442bb3c3e5d69ef58b44b5383ad7cf3bc504f004e1718e40","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-06-28T17:59:33Z","title_canon_sha256":"96dcd19035b316638979c7a18193c6ab4c5b0a9dbb5fa6b789f962ac2795b44a"},"schema_version":"1.0","source":{"id":"2106.14881","kind":"arxiv","version":3}},"canonical_sha256":"ba287fb322e43964b0d3c613a01f908926359a657be37a7c719619436edaf81a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ba287fb322e43964b0d3c613a01f908926359a657be37a7c719619436edaf81a","first_computed_at":"2026-07-05T03:25:32.496944Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T03:25:32.496944Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"01v487iPDF/r3iNBWuFK5O5uaScTEGzBPb/pTGLrLvEzLXCXjiK1hG0TJqlEUDUtIogfnny58PN8COMyjFbeBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T03:25:32.497340Z","signed_message":"canonical_sha256_bytes"},"source_id":"2106.14881","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a8b512a08b1b8b47e6b42fe4cd6da607b69ce8a9ea1ada35a22d7ddca2477eb0","sha256:7a13ba1a2da31c2e0e9c3a4cb1fcde89e17037ce0518d64a21d8b9a58f94dc3b"],"state_sha256":"1369a5028a105a783adbed9868155252824398ef7523a8b5192c228e22631989"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"v7lfV0ALOiL55viJxaZ6ndyXA84+rU8qrDFbjuv8llk7dhLqYPG+t659Qxxxsrko8bzQixmAfCAOUr8/J4w1Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T00:35:31.927077Z","bundle_sha256":"c51defa1a4bf4f26872cb8ed437acc7610204f73b02af162d16c763eb9ed1181"}}