{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:QD7ZLHHRYH3CBUJQBRPTTNUUQZ","short_pith_number":"pith:QD7ZLHHR","schema_version":"1.0","canonical_sha256":"80ff959cf1c1f620d1300c5f39b69486704d9a5a06df16a09b0155299c883441","source":{"kind":"arxiv","id":"2001.06782","version":4},"attestation_state":"computed","paper":{"title":"Gradient Surgery for Multi-Task Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.RO","stat.ML"],"primary_cat":"cs.LG","authors_text":"Abhishek Gupta, Chelsea Finn, Karol Hausman, Saurabh Kumar, Sergey Levine, Tianhe Yu","submitted_at":"2020-01-19T06:33:47Z","abstract_excerpt":"While deep learning and deep reinforcement learning (RL) systems have demonstrated impressive results in domains such as image classification, game playing, and robotic control, data efficiency remains a major challenge. Multi-task learning has emerged as a promising approach for sharing structure across multiple tasks to enable more efficient learning. However, the multi-task setting presents a number of optimization challenges, making it difficult to realize large efficiency gains compared to learning tasks independently. The reasons why multi-task learning is so challenging compared to sing"},"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":"2001.06782","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-01-19T06:33:47Z","cross_cats_sorted":["cs.CV","cs.RO","stat.ML"],"title_canon_sha256":"9fe1b0af9ab8c966a073c5cfb489e43ccde5edf3939195fcda3854bdab08db14","abstract_canon_sha256":"bd3fcbfcdc190e196ae68d3229b62e7395362adde695eba150e11c10d501cc07"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:01:10.468458Z","signature_b64":"GyZYlll/766Alf1MXHU56/t2yRgZQNFPdCKut4ZJNTfzd5+V8cArs3fM4KFkbFG8lp9PVnGrpnHIFA27DeArAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"80ff959cf1c1f620d1300c5f39b69486704d9a5a06df16a09b0155299c883441","last_reissued_at":"2026-07-05T02:01:10.467957Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:01:10.467957Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradient Surgery for Multi-Task Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.RO","stat.ML"],"primary_cat":"cs.LG","authors_text":"Abhishek Gupta, Chelsea Finn, Karol Hausman, Saurabh Kumar, Sergey Levine, Tianhe Yu","submitted_at":"2020-01-19T06:33:47Z","abstract_excerpt":"While deep learning and deep reinforcement learning (RL) systems have demonstrated impressive results in domains such as image classification, game playing, and robotic control, data efficiency remains a major challenge. Multi-task learning has emerged as a promising approach for sharing structure across multiple tasks to enable more efficient learning. However, the multi-task setting presents a number of optimization challenges, making it difficult to realize large efficiency gains compared to learning tasks independently. The reasons why multi-task learning is so challenging compared to sing"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2001.06782","kind":"arxiv","version":4},"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/2001.06782/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":"2001.06782","created_at":"2026-07-05T02:01:10.468009+00:00"},{"alias_kind":"arxiv_version","alias_value":"2001.06782v4","created_at":"2026-07-05T02:01:10.468009+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2001.06782","created_at":"2026-07-05T02:01:10.468009+00:00"},{"alias_kind":"pith_short_12","alias_value":"QD7ZLHHRYH3C","created_at":"2026-07-05T02:01:10.468009+00:00"},{"alias_kind":"pith_short_16","alias_value":"QD7ZLHHRYH3CBUJQ","created_at":"2026-07-05T02:01:10.468009+00:00"},{"alias_kind":"pith_short_8","alias_value":"QD7ZLHHR","created_at":"2026-07-05T02:01:10.468009+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20272","citing_title":"Efficiently Linking Real Scenes with Synthetic Data Generation for AI-based Cognitive Robotics and Computer Vision Applications","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20344","citing_title":"Quantum ring all-reduce: communication and privacy advantages for distributed learning","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10929","citing_title":"Recoverable but Not Stationary:Local Linear Structures in Weights and Activations","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05906","citing_title":"ACE-SQL: Adaptive Co-Optimization via Empirical Credit Assignment for Text-to-SQL","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05800","citing_title":"SALT: When More Rollouts Don't Help in Group-Based Policy Optimization and How to Make Them Matter","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02381","citing_title":"A Mathematical Conflict Framework for Contextual Data Modulation","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25025","citing_title":"Micro-Swarm Locomotion Optimization in Dynamic Flow using Multi-Objective Multi-Agent Reinforcement Learning","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30011","citing_title":"T3R: Deeper Test-Time Adaptation for Graph Neural Networks via Gradient Rotation","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15394","citing_title":"Representation Without Reward: A JEPA Audit for LLM Fine-Tuning","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2109.13396","citing_title":"Bridge Data: Boosting Generalization of Robotic Skills with Cross-Domain Datasets","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01224","citing_title":"Lost in the Tower of Babel: The Adverse Effects of Incidental Multilingualism in LLMs","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20720","citing_title":"COMPASS: COntinual Multilingual PEFT with Adaptive Semantic Sampling","ref_index":83,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ","json":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ.json","graph_json":"https://pith.science/api/pith-number/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/graph.json","events_json":"https://pith.science/api/pith-number/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/events.json","paper":"https://pith.science/paper/QD7ZLHHR"},"agent_actions":{"view_html":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ","download_json":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ.json","view_paper":"https://pith.science/paper/QD7ZLHHR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2001.06782&json=true","fetch_graph":"https://pith.science/api/pith-number/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/graph.json","fetch_events":"https://pith.science/api/pith-number/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/action/storage_attestation","attest_author":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/action/author_attestation","sign_citation":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/action/citation_signature","submit_replication":"https://pith.science/pith/QD7ZLHHRYH3CBUJQBRPTTNUUQZ/action/replication_record"}},"created_at":"2026-07-05T02:01:10.468009+00:00","updated_at":"2026-07-05T02:01:10.468009+00:00"}