{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:EYYQJAJVAQZHPP2XL45FD2YJZ5","short_pith_number":"pith:EYYQJAJV","schema_version":"1.0","canonical_sha256":"2631048135043277bf575f3a51eb09cf5b2a12e16feb6cea208de211ecda7251","source":{"kind":"arxiv","id":"1611.03673","version":3},"attestation_state":"computed","paper":{"title":"Learning to Navigate in Complex Environments","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG","cs.RO"],"primary_cat":"cs.AI","authors_text":"Andrea Banino, Andrew J. Ballard, Dharshan Kumaran, Fabio Viola, Hubert Soyer, Koray Kavukcuoglu, Laurent Sifre, Misha Denil, Piotr Mirowski, Raia Hadsell, Razvan Pascanu, Ross Goroshin","submitted_at":"2016-11-11T12:14:45Z","abstract_excerpt":"Learning to navigate in complex environments with dynamic elements is an important milestone in developing AI agents. In this work we formulate the navigation question as a reinforcement learning problem and show that data efficiency and task performance can be dramatically improved by relying on additional auxiliary tasks leveraging multimodal sensory inputs. In particular we consider jointly learning the goal-driven reinforcement learning problem with auxiliary depth prediction and loop closure classification tasks. This approach can learn to navigate from raw sensory input in complicated 3D"},"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":"1611.03673","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2016-11-11T12:14:45Z","cross_cats_sorted":["cs.CV","cs.LG","cs.RO"],"title_canon_sha256":"0176ecb7fc5f1683ab0620ca778950c593aa24832d59960173fa2e8829003983","abstract_canon_sha256":"3b189e79d9fa6177f7369d367402439a8ed8942496b4c57497bda01f412769c4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:52:54.972846Z","signature_b64":"/cEeiUXFlM3yTTuqiHiB0Mm2aIBgkObXzUgrPfFxqLI5Zjo08aM3rDsMihAK04+q0tUhUj1Wtrhi3FK8Ia+rAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2631048135043277bf575f3a51eb09cf5b2a12e16feb6cea208de211ecda7251","last_reissued_at":"2026-05-18T00:52:54.972293Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:52:54.972293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning to Navigate in Complex Environments","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG","cs.RO"],"primary_cat":"cs.AI","authors_text":"Andrea Banino, Andrew J. Ballard, Dharshan Kumaran, Fabio Viola, Hubert Soyer, Koray Kavukcuoglu, Laurent Sifre, Misha Denil, Piotr Mirowski, Raia Hadsell, Razvan Pascanu, Ross Goroshin","submitted_at":"2016-11-11T12:14:45Z","abstract_excerpt":"Learning to navigate in complex environments with dynamic elements is an important milestone in developing AI agents. In this work we formulate the navigation question as a reinforcement learning problem and show that data efficiency and task performance can be dramatically improved by relying on additional auxiliary tasks leveraging multimodal sensory inputs. In particular we consider jointly learning the goal-driven reinforcement learning problem with auxiliary depth prediction and loop closure classification tasks. This approach can learn to navigate from raw sensory input in complicated 3D"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1611.03673","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":""},"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":"1611.03673","created_at":"2026-05-18T00:52:54.972400+00:00"},{"alias_kind":"arxiv_version","alias_value":"1611.03673v3","created_at":"2026-05-18T00:52:54.972400+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1611.03673","created_at":"2026-05-18T00:52:54.972400+00:00"},{"alias_kind":"pith_short_12","alias_value":"EYYQJAJVAQZH","created_at":"2026-05-18T12:30:15.759754+00:00"},{"alias_kind":"pith_short_16","alias_value":"EYYQJAJVAQZHPP2X","created_at":"2026-05-18T12:30:15.759754+00:00"},{"alias_kind":"pith_short_8","alias_value":"EYYQJAJV","created_at":"2026-05-18T12:30:15.759754+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":6,"sample":[{"citing_arxiv_id":"1906.09237","citing_title":"Shaping Belief States with Generative Environment Models for RL","ref_index":51,"is_internal_anchor":true},{"citing_arxiv_id":"1906.09510","citing_title":"Learning Belief Representations for Imitation Learning in POMDPs","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"1906.11951","citing_title":"Supervise Thyself: Examining Self-Supervised Representations in Interactive Environments","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23551","citing_title":"Goal-Conditioned Agents that Learn Everything All at Once","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22814","citing_title":"Remember to be Curious: Episodic Context and Persistent Worlds for 3D Exploration","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2510.09574","citing_title":"Online Structure Learning and Planning for Autonomous Robot Navigation using Active Inference","ref_index":25,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5","json":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5.json","graph_json":"https://pith.science/api/pith-number/EYYQJAJVAQZHPP2XL45FD2YJZ5/graph.json","events_json":"https://pith.science/api/pith-number/EYYQJAJVAQZHPP2XL45FD2YJZ5/events.json","paper":"https://pith.science/paper/EYYQJAJV"},"agent_actions":{"view_html":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5","download_json":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5.json","view_paper":"https://pith.science/paper/EYYQJAJV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1611.03673&json=true","fetch_graph":"https://pith.science/api/pith-number/EYYQJAJVAQZHPP2XL45FD2YJZ5/graph.json","fetch_events":"https://pith.science/api/pith-number/EYYQJAJVAQZHPP2XL45FD2YJZ5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5/action/storage_attestation","attest_author":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5/action/author_attestation","sign_citation":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5/action/citation_signature","submit_replication":"https://pith.science/pith/EYYQJAJVAQZHPP2XL45FD2YJZ5/action/replication_record"}},"created_at":"2026-05-18T00:52:54.972400+00:00","updated_at":"2026-05-18T00:52:54.972400+00:00"}