{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2015:WZ25MAV37OTHZQQPWI6Z6B5CAD","short_pith_number":"pith:WZ25MAV3","schema_version":"1.0","canonical_sha256":"b675d602bbfba67cc20fb23d9f07a200da75e2e166c27742a5a3bd56719212ba","source":{"kind":"arxiv","id":"1511.02580","version":1},"attestation_state":"computed","paper":{"title":"How far can we go without convolution: Improving fully-connected networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE"],"primary_cat":"cs.LG","authors_text":"Kishore Konda, Roland Memisevic, Zhouhan Lin","submitted_at":"2015-11-09T06:56:24Z","abstract_excerpt":"We propose ways to improve the performance of fully connected networks. We found that two approaches in particular have a strong effect on performance: linear bottleneck layers and unsupervised pre-training using autoencoders without hidden unit biases. We show how both approaches can be related to improving gradient flow and reducing sparsity in the network. We show that a fully connected network can yield approximately 70% classification accuracy on the permutation-invariant CIFAR-10 task, which is much higher than the current state-of-the-art. By adding deformations to the training data, th"},"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":"1511.02580","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2015-11-09T06:56:24Z","cross_cats_sorted":["cs.NE"],"title_canon_sha256":"144b06c6cd33b7a89e184c0f759068beb54d0b58b1bc8134f3dd175ed9c41ad8","abstract_canon_sha256":"745f501a272eed0a41d0627e5581500fba112f66a2310650a559a45133ee0471"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:27:29.911715Z","signature_b64":"Y8tgau8ZLK3Y1Es3THCpcj2tCjWI2G0Zn/QVYk3SGSR0CURHUybPdHqRSLCmZCraKzUXyQw9bpM2zbTMewhPBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b675d602bbfba67cc20fb23d9f07a200da75e2e166c27742a5a3bd56719212ba","last_reissued_at":"2026-05-18T01:27:29.911142Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:27:29.911142Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How far can we go without convolution: Improving fully-connected networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE"],"primary_cat":"cs.LG","authors_text":"Kishore Konda, Roland Memisevic, Zhouhan Lin","submitted_at":"2015-11-09T06:56:24Z","abstract_excerpt":"We propose ways to improve the performance of fully connected networks. We found that two approaches in particular have a strong effect on performance: linear bottleneck layers and unsupervised pre-training using autoencoders without hidden unit biases. We show how both approaches can be related to improving gradient flow and reducing sparsity in the network. We show that a fully connected network can yield approximately 70% classification accuracy on the permutation-invariant CIFAR-10 task, which is much higher than the current state-of-the-art. By adding deformations to the training data, th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1511.02580","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":""},"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":"1511.02580","created_at":"2026-05-18T01:27:29.911210+00:00"},{"alias_kind":"arxiv_version","alias_value":"1511.02580v1","created_at":"2026-05-18T01:27:29.911210+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1511.02580","created_at":"2026-05-18T01:27:29.911210+00:00"},{"alias_kind":"pith_short_12","alias_value":"WZ25MAV37OTH","created_at":"2026-05-18T12:29:47.479230+00:00"},{"alias_kind":"pith_short_16","alias_value":"WZ25MAV37OTHZQQP","created_at":"2026-05-18T12:29:47.479230+00:00"},{"alias_kind":"pith_short_8","alias_value":"WZ25MAV3","created_at":"2026-05-18T12:29:47.479230+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD","json":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD.json","graph_json":"https://pith.science/api/pith-number/WZ25MAV37OTHZQQPWI6Z6B5CAD/graph.json","events_json":"https://pith.science/api/pith-number/WZ25MAV37OTHZQQPWI6Z6B5CAD/events.json","paper":"https://pith.science/paper/WZ25MAV3"},"agent_actions":{"view_html":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD","download_json":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD.json","view_paper":"https://pith.science/paper/WZ25MAV3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1511.02580&json=true","fetch_graph":"https://pith.science/api/pith-number/WZ25MAV37OTHZQQPWI6Z6B5CAD/graph.json","fetch_events":"https://pith.science/api/pith-number/WZ25MAV37OTHZQQPWI6Z6B5CAD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD/action/storage_attestation","attest_author":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD/action/author_attestation","sign_citation":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD/action/citation_signature","submit_replication":"https://pith.science/pith/WZ25MAV37OTHZQQPWI6Z6B5CAD/action/replication_record"}},"created_at":"2026-05-18T01:27:29.911210+00:00","updated_at":"2026-05-18T01:27:29.911210+00:00"}