{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:HFZ4AW4ZBUX3UKDEVI7ORDUBWM","short_pith_number":"pith:HFZ4AW4Z","schema_version":"1.0","canonical_sha256":"3973c05b990d2fba2864aa3ee88e81b33b9d45cc742cc173f8b245bad3a676a3","source":{"kind":"arxiv","id":"2105.03404","version":2},"attestation_state":"computed","paper":{"title":"ResMLP: Feedforward networks for image classification with data-efficient training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alaaeldin El-Nouby, Armand Joulin, Edouard Grave, Gabriel Synnaeve, Gautier Izacard, Herv\\'e J\\'egou, Hugo Touvron, Jakob Verbeek, Mathilde Caron, Matthieu Cord, Piotr Bojanowski","submitted_at":"2021-05-07T17:31:44Z","abstract_excerpt":"We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification. It is a simple residual network that alternates (i) a linear layer in which image patches interact, independently and identically across channels, and (ii) a two-layer feed-forward network in which channels interact independently per patch. When trained with a modern training strategy using heavy data-augmentation and optionally distillation, it attains surprisingly good accuracy/complexity trade-offs on ImageNet. We also train ResMLP models in a self-supervised setup, to further remove pri"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2105.03404","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-05-07T17:31:44Z","cross_cats_sorted":[],"title_canon_sha256":"bd344a9589bda026a2158350d02673bd860e4e6624dea5d35ddb6935865327b4","abstract_canon_sha256":"8fc7dbdf81d84c5302ece60da867ab4b0ce13091c81dbbadc05a2ebe9859ab6f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:48:11.185278Z","signature_b64":"UGkwkwexiakXEFh1YYW5YdoHz8XO78IYuHXS0OOeZ5D8nNrbYt5yE5EzukjGHfulr8POUK+KHdyptF7AYdUXAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3973c05b990d2fba2864aa3ee88e81b33b9d45cc742cc173f8b245bad3a676a3","last_reissued_at":"2026-07-05T02:48:11.184808Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:48:11.184808Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ResMLP: Feedforward networks for image classification with data-efficient training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alaaeldin El-Nouby, Armand Joulin, Edouard Grave, Gabriel Synnaeve, Gautier Izacard, Herv\\'e J\\'egou, Hugo Touvron, Jakob Verbeek, Mathilde Caron, Matthieu Cord, Piotr Bojanowski","submitted_at":"2021-05-07T17:31:44Z","abstract_excerpt":"We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification. It is a simple residual network that alternates (i) a linear layer in which image patches interact, independently and identically across channels, and (ii) a two-layer feed-forward network in which channels interact independently per patch. When trained with a modern training strategy using heavy data-augmentation and optionally distillation, it attains surprisingly good accuracy/complexity trade-offs on ImageNet. We also train ResMLP models in a self-supervised setup, to further remove pri"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2105.03404","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/2105.03404/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2105.03404","created_at":"2026-07-05T02:48:11.184863+00:00"},{"alias_kind":"arxiv_version","alias_value":"2105.03404v2","created_at":"2026-07-05T02:48:11.184863+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2105.03404","created_at":"2026-07-05T02:48:11.184863+00:00"},{"alias_kind":"pith_short_12","alias_value":"HFZ4AW4ZBUX3","created_at":"2026-07-05T02:48:11.184863+00:00"},{"alias_kind":"pith_short_16","alias_value":"HFZ4AW4ZBUX3UKDE","created_at":"2026-07-05T02:48:11.184863+00:00"},{"alias_kind":"pith_short_8","alias_value":"HFZ4AW4Z","created_at":"2026-07-05T02:48:11.184863+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25986","citing_title":"The Inference-Compute Frontier and a Latency-Efficient Architecture for Limit Order Book Prediction","ref_index":29,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM","json":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM.json","graph_json":"https://pith.science/api/pith-number/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/graph.json","events_json":"https://pith.science/api/pith-number/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/events.json","paper":"https://pith.science/paper/HFZ4AW4Z"},"agent_actions":{"view_html":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM","download_json":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM.json","view_paper":"https://pith.science/paper/HFZ4AW4Z","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2105.03404&json=true","fetch_graph":"https://pith.science/api/pith-number/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/graph.json","fetch_events":"https://pith.science/api/pith-number/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/action/storage_attestation","attest_author":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/action/author_attestation","sign_citation":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/action/citation_signature","submit_replication":"https://pith.science/pith/HFZ4AW4ZBUX3UKDEVI7ORDUBWM/action/replication_record"}},"created_at":"2026-07-05T02:48:11.184863+00:00","updated_at":"2026-07-05T02:48:11.184863+00:00"}