{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:77IZAOLOPNEU4VOASX77YJGLP4","short_pith_number":"pith:77IZAOLO","schema_version":"1.0","canonical_sha256":"ffd190396e7b494e55c095fffc24cb7f330f218abae274898842ba80e838122d","source":{"kind":"arxiv","id":"2111.05803","version":2},"attestation_state":"computed","paper":{"title":"Gradients are Not All You Need","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"C. Daniel Freeman, Luke Metz, Samuel S. Schoenholz, Tal Kachman","submitted_at":"2021-11-10T16:51:04Z","abstract_excerpt":"Differentiable programming techniques are widely used in the community and are responsible for the machine learning renaissance of the past several decades. While these methods are powerful, they have limits. In this short report, we discuss a common chaos based failure mode which appears in a variety of differentiable circumstances, ranging from recurrent neural networks and numerical physics simulation to training learned optimizers. We trace this failure to the spectrum of the Jacobian of the system under study, and provide criteria for when a practitioner might expect this failure to spoil"},"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":"2111.05803","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-11-10T16:51:04Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"27bbc85190ec747b388476de29a13cad87858518afbb5eb96ca38abec05b97e5","abstract_canon_sha256":"d5a459df5ec2b3945e2baf46dd2a0a3bb59983eac9bbd3ae76677c9b31ca794a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:50:26.302102Z","signature_b64":"3xfJ7TMAS8JygSPPSbHHcOUpPatJapwdUWUCUcTN9A7t2gY7gKL+7JjRxcO/l0p/W9sg9ytbsNlN3qyTKwj0Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ffd190396e7b494e55c095fffc24cb7f330f218abae274898842ba80e838122d","last_reissued_at":"2026-07-05T03:50:26.301661Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:50:26.301661Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradients are Not All You Need","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"C. Daniel Freeman, Luke Metz, Samuel S. Schoenholz, Tal Kachman","submitted_at":"2021-11-10T16:51:04Z","abstract_excerpt":"Differentiable programming techniques are widely used in the community and are responsible for the machine learning renaissance of the past several decades. While these methods are powerful, they have limits. In this short report, we discuss a common chaos based failure mode which appears in a variety of differentiable circumstances, ranging from recurrent neural networks and numerical physics simulation to training learned optimizers. We trace this failure to the spectrum of the Jacobian of the system under study, and provide criteria for when a practitioner might expect this failure to spoil"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.05803","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/2111.05803/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":"2111.05803","created_at":"2026-07-05T03:50:26.301718+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.05803v2","created_at":"2026-07-05T03:50:26.301718+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.05803","created_at":"2026-07-05T03:50:26.301718+00:00"},{"alias_kind":"pith_short_12","alias_value":"77IZAOLOPNEU","created_at":"2026-07-05T03:50:26.301718+00:00"},{"alias_kind":"pith_short_16","alias_value":"77IZAOLOPNEU4VOA","created_at":"2026-07-05T03:50:26.301718+00:00"},{"alias_kind":"pith_short_8","alias_value":"77IZAOLO","created_at":"2026-07-05T03:50:26.301718+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25761","citing_title":"Bridging Spherical Black-Box Optimizers","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21866","citing_title":"SurGE: Surrogate Gradient-guided Evolution for Co-design of Legged Robots with Parallel Elasticity","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00215","citing_title":"ELMP: Efficient Learning for Motion Planning via Analytical Policy Gradients","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04880","citing_title":"MAOAM: Unified Object and Material Selection with Vision-Language Models","ref_index":92,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26478","citing_title":"Efficient On-policy Visual-RL via Stochastic Decoupled Policy Gradient","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2506.01167","citing_title":"Accelerated Learning with Linear Temporal Logic using Differentiable Simulation","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2604.00904","citing_title":"Fatigue-Aware Learning to Defer via Constrained Optimisation","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02779","citing_title":"Vision-Based End-to-End Learning for UAV Traversal of Irregular Gaps via Differentiable Simulation","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07979","citing_title":"Differentiable hybrid force fields support scalable autonomous electrolyte discovery","ref_index":52,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4","json":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4.json","graph_json":"https://pith.science/api/pith-number/77IZAOLOPNEU4VOASX77YJGLP4/graph.json","events_json":"https://pith.science/api/pith-number/77IZAOLOPNEU4VOASX77YJGLP4/events.json","paper":"https://pith.science/paper/77IZAOLO"},"agent_actions":{"view_html":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4","download_json":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4.json","view_paper":"https://pith.science/paper/77IZAOLO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.05803&json=true","fetch_graph":"https://pith.science/api/pith-number/77IZAOLOPNEU4VOASX77YJGLP4/graph.json","fetch_events":"https://pith.science/api/pith-number/77IZAOLOPNEU4VOASX77YJGLP4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4/action/storage_attestation","attest_author":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4/action/author_attestation","sign_citation":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4/action/citation_signature","submit_replication":"https://pith.science/pith/77IZAOLOPNEU4VOASX77YJGLP4/action/replication_record"}},"created_at":"2026-07-05T03:50:26.301718+00:00","updated_at":"2026-07-05T03:50:26.301718+00:00"}