{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2015:VZBVR22VSJQJW4LQV25FUKMBD5","short_pith_number":"pith:VZBVR22V","schema_version":"1.0","canonical_sha256":"ae4358eb5592609b7170aeba5a29811f74029029eed9e444aec2dbfb9538e3b1","source":{"kind":"arxiv","id":"1511.06297","version":2},"attestation_state":"computed","paper":{"title":"Conditional Computation in Neural Networks for faster models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Doina Precup, Emmanuel Bengio, Joelle Pineau, Pierre-Luc Bacon","submitted_at":"2015-11-19T18:40:22Z","abstract_excerpt":"Deep learning has become the state-of-art tool in many applications, but the evaluation and training of deep models can be time-consuming and computationally expensive. The conditional computation approach has been proposed to tackle this problem (Bengio et al., 2013; Davis & Arel, 2013). It operates by selectively activating only parts of the network at a time. In this paper, we use reinforcement learning as a tool to optimize conditional computation policies. More specifically, we cast the problem of learning activation-dependent policies for dropping out blocks of units as a reinforcement l"},"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":"1511.06297","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2015-11-19T18:40:22Z","cross_cats_sorted":[],"title_canon_sha256":"dc43c7cd717cb91a9a339dfb501c13d7cda1b69e80a2222fead2fae46548f1db","abstract_canon_sha256":"7485a9f9afd9790b3aa6cf1f60fa27ca9d9069a60e743b0bb0ab3d17c8433419"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:23:11.828224Z","signature_b64":"yV/b+YTa3If61oHWVF6rbVgdnQN0nmY++ShL42eMGaLhDuRiKwN+LPfV1YuomOVG1c5b8IScb1HjV5Qg2o8gAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ae4358eb5592609b7170aeba5a29811f74029029eed9e444aec2dbfb9538e3b1","last_reissued_at":"2026-05-18T01:23:11.827646Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:23:11.827646Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Conditional Computation in Neural Networks for faster models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Doina Precup, Emmanuel Bengio, Joelle Pineau, Pierre-Luc Bacon","submitted_at":"2015-11-19T18:40:22Z","abstract_excerpt":"Deep learning has become the state-of-art tool in many applications, but the evaluation and training of deep models can be time-consuming and computationally expensive. The conditional computation approach has been proposed to tackle this problem (Bengio et al., 2013; Davis & Arel, 2013). It operates by selectively activating only parts of the network at a time. In this paper, we use reinforcement learning as a tool to optimize conditional computation policies. More specifically, we cast the problem of learning activation-dependent policies for dropping out blocks of units as a reinforcement l"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1511.06297","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":""},"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":"1511.06297","created_at":"2026-05-18T01:23:11.827735+00:00"},{"alias_kind":"arxiv_version","alias_value":"1511.06297v2","created_at":"2026-05-18T01:23:11.827735+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1511.06297","created_at":"2026-05-18T01:23:11.827735+00:00"},{"alias_kind":"pith_short_12","alias_value":"VZBVR22VSJQJ","created_at":"2026-05-18T12:29:47.479230+00:00"},{"alias_kind":"pith_short_16","alias_value":"VZBVR22VSJQJW4LQ","created_at":"2026-05-18T12:29:47.479230+00:00"},{"alias_kind":"pith_short_8","alias_value":"VZBVR22V","created_at":"2026-05-18T12:29:47.479230+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2502.15315","citing_title":"Tight Clusters Make Specialized Experts","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2502.05564","citing_title":"TabICL: A Tabular Foundation Model for In-Context Learning on Large Data","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"1603.08983","citing_title":"Adaptive Computation Time for Recurrent Neural Networks","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09630","citing_title":"Scratchpad Patching: Decoupling Compute from Patch Size in Byte-Level Language Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07160","citing_title":"TENNOR: Trustworthy Execution for Neural Networks through Obliviousness and Retrievals","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13546","citing_title":"Learning Inference Concurrency in DynamicGate MLP Structural and Mathematical Justification","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14853","citing_title":"Adaptive Test-Time Compute Allocation for Reasoning LLMs via Constrained Policy Optimization","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"1701.06538","citing_title":"Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5","json":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5.json","graph_json":"https://pith.science/api/pith-number/VZBVR22VSJQJW4LQV25FUKMBD5/graph.json","events_json":"https://pith.science/api/pith-number/VZBVR22VSJQJW4LQV25FUKMBD5/events.json","paper":"https://pith.science/paper/VZBVR22V"},"agent_actions":{"view_html":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5","download_json":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5.json","view_paper":"https://pith.science/paper/VZBVR22V","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1511.06297&json=true","fetch_graph":"https://pith.science/api/pith-number/VZBVR22VSJQJW4LQV25FUKMBD5/graph.json","fetch_events":"https://pith.science/api/pith-number/VZBVR22VSJQJW4LQV25FUKMBD5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5/action/storage_attestation","attest_author":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5/action/author_attestation","sign_citation":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5/action/citation_signature","submit_replication":"https://pith.science/pith/VZBVR22VSJQJW4LQV25FUKMBD5/action/replication_record"}},"created_at":"2026-05-18T01:23:11.827735+00:00","updated_at":"2026-05-18T01:23:11.827735+00:00"}