{"as_of":"2026-08-16T13:13:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:67b4647b561c963be089f6bf2f9001d52c556261d725b50a6724f326ffa3f064","coverage":[{"denominator":0,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":5,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":5,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-16T06:30:59.297886+00:00","state":"measured"},{"denominator":5,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":5,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-15T23:43:39.672597Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"pith","source_observed_at":"2026-08-14T10:36:30.762220Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"1710.07406","last_updated":"2017-10-20T03:34:56Z","snapshot_observed_at":"2026-08-14T20:21:59.319940Z","submitted_at":"2017-10-20T03:34:56Z","title":"First-order Methods Almost Always Avoid Saddle Points","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1710.07406","snapshot_observed_at":"2026-08-14T15:14:53.411614Z","title":"First-order methods almost always avoid sa ddle points","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"1908.01753","last_updated":"2019-08-05T17:50:55Z","snapshot_observed_at":"2026-08-15T21:56:44.385213Z","submitted_at":"2019-08-05T17:50:55Z","title":"Extending the step-size restriction for gradient descent to avoid strict saddle points","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-14T15:14:53.411614Z"},"links":{"cited_paper":"/paper/1710.07406","citing_paper":"/paper/1908.01753"},"observation_digest":"sha256:1d4ed00381442f4577dcc58878e1d92e8776d38d79bc73a93c17a74d4ad2d25c","observation_id":"50aaf26f-96d6-42ba-a530-eb26bbd29da0","resolution":{"observed_at":"2026-08-14T15:14:53.411614Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1710.07406","last_updated":"2017-10-20T03:34:56Z","snapshot_observed_at":"2026-08-14T20:21:59.319940Z","submitted_at":"2017-10-20T03:34:56Z","title":"First-order Methods Almost Always Avoid Saddle Points","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1710.07406","snapshot_observed_at":"2026-08-14T12:47:24.833526Z","title":"First-order methods almost always avoid saddle p oints,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"1908.06693","last_updated":"2019-08-19T11:08:37Z","snapshot_observed_at":"2026-08-14T12:34:49.345336Z","submitted_at":"2019-08-19T11:08:37Z","title":"Distributed Stochastic Gradient Method for Non-Convex Problems with Applications in Supervised Learning","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-14T12:47:24.833526Z"},"links":{"cited_paper":"/paper/1710.07406","citing_paper":"/paper/1908.06693"},"observation_digest":"sha256:18fccb4386def8812833e00442752642359cc8448e9181bd24a9fb98b73e55bb","observation_id":"27484ef9-3f47-4fb7-8b89-b2a93a6833dc","resolution":{"observed_at":"2026-08-14T12:47:24.833526Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1710.07406","last_updated":"2017-10-20T03:34:56Z","snapshot_observed_at":"2026-08-14T20:21:59.319940Z","submitted_at":"2017-10-20T03:34:56Z","title":"First-order Methods Almost Always Avoid Saddle Points","version":1},"cited_work":{"arxiv_id":"1710.07406","doi":null,"metadata_source":"pith","pith_arxiv_id":"1710.07406","snapshot_observed_at":"2026-08-14T10:36:30.762220Z","title":"First-order Methods Almost Always Avoid Saddle Points","venue":"stat.ML","work_id":"ccc07496-2912-4402-94dd-360380324f3d","year":2017},"citing_paper":{"arxiv_id":"1908.10920","last_updated":"2019-09-28T20:40:02Z","snapshot_observed_at":"2026-08-15T12:34:04.854283Z","submitted_at":"2019-08-28T19:36:23Z","title":"Deep Learning Theory Review: An Optimal Control and Dynamical Systems Perspective","version":2},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-14T10:36:29.629807Z"},"links":{"cited_paper":"/paper/1710.07406","citing_paper":"/paper/1908.10920"},"observation_digest":"sha256:201f001ade3f0c10c23f125211d4b8585a92503d7e27163f05ff252a10704cd6","observation_id":"207f2e1d-7f74-4a91-b4c6-7071eda18cd5","resolution":{"observed_at":"2026-08-14T10:36:30.766648Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1710.07406","last_updated":"2017-10-20T03:34:56Z","snapshot_observed_at":"2026-08-14T20:21:59.319940Z","submitted_at":"2017-10-20T03:34:56Z","title":"First-order Methods Almost Always Avoid Saddle Points","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1710.07406","snapshot_observed_at":"2026-08-15T23:43:39.672597Z","title":"First-order methods almost always avoid saddle points.arXiv preprint arXiv:1710.07406, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2505.04037","last_updated":"2025-05-07T00:33:14Z","snapshot_observed_at":"2026-08-15T23:37:05.369877Z","submitted_at":"2025-05-07T00:33:14Z","title":"Learning based convex approximation for constrained parametric optimization","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-15T23:43:39.672597Z"},"links":{"cited_paper":"/paper/1710.07406","citing_paper":"/paper/2505.04037"},"observation_digest":"sha256:bc10a95f493fa56de0a144908703164ed2c66e2508d7a73cf83baa428096c40b","observation_id":"24952eca-8571-4f68-ba10-c84e72b6ed57","resolution":{"observed_at":"2026-08-15T23:43:39.672597Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1710.07406","last_updated":"2017-10-20T03:34:56Z","snapshot_observed_at":"2026-08-14T20:21:59.319940Z","submitted_at":"2017-10-20T03:34:56Z","title":"First-order Methods Almost Always Avoid Saddle Points","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1710.07406","snapshot_observed_at":"2026-08-15T20:26:51.797163Z","title":null,"venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2505.13582","last_updated":"2025-05-19T15:50:36Z","snapshot_observed_at":"2026-08-15T20:14:12.089141Z","submitted_at":"2025-05-19T15:50:36Z","title":"Uncovering Critical Sets of Deep Neural Networks via Sample-Independent Critical Lifting","version":1},"reference_index":11,"source":"arxiv_source","source_observed_at":"2026-08-15T20:26:51.797163Z"},"links":{"cited_paper":"/paper/1710.07406","citing_paper":"/paper/2505.13582"},"observation_digest":"sha256:d00c68f0ae9f9eb0a3f26c8ccb60324632587d13791fcc66ffc378d40c04ac90","observation_id":"aff3332f-ccb3-49cc-ad0a-c49c1bf8b27d","resolution":{"observed_at":"2026-08-15T20:26:51.797163Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"links":{"evidence":"/evidence","html":"/paper/1710.07406/citation-record","integrity":"/paper/1710.07406/integrity","json":"/paper/1710.07406/citation-record.json","paper":"/paper/1710.07406"},"outbound":[],"paper":{"arxiv_id":"1710.07406","last_updated":"2017-10-20T03:34:56Z","latest_version":1,"primary_category":"stat.ML","snapshot_observed_at":"2026-08-14T20:21:59.319940Z","submitted_at":"2017-10-20T03:34:56Z","title":"First-order Methods Almost Always Avoid Saddle Points"},"reference_resolution":{"displayed":0,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":0,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":0},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"thesis":"As of 16 August 2026, this Paper Citation Record lists 0 of 0 outbound references and 5 inbound Pith citation observations for arXiv:1710.07406."}