{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2016:JKTQYFNEHNRFYFZR7DN6FG67LW","short_pith_number":"pith:JKTQYFNE","canonical_record":{"source":{"id":"1612.04010","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2016-12-13T03:27:11Z","cross_cats_sorted":[],"title_canon_sha256":"a0c67bc14eae1774808058c00fefa9243d47863d3102a5f2f8b864275d6b5698","abstract_canon_sha256":"aad78584267f1d4319df830536f49ae22e8b7eb39da6ecbe19e4f6867a0fc6d2"},"schema_version":"1.0"},"canonical_sha256":"4aa70c15a43b625c1731f8dbe29bdf5d99a0bcb3212511e86210431de7c6aa5c","source":{"kind":"arxiv","id":"1612.04010","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1612.04010","created_at":"2026-05-18T00:28:32Z"},{"alias_kind":"arxiv_version","alias_value":"1612.04010v4","created_at":"2026-05-18T00:28:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1612.04010","created_at":"2026-05-18T00:28:32Z"},{"alias_kind":"pith_short_12","alias_value":"JKTQYFNEHNRF","created_at":"2026-05-18T12:30:25Z"},{"alias_kind":"pith_short_16","alias_value":"JKTQYFNEHNRFYFZR","created_at":"2026-05-18T12:30:25Z"},{"alias_kind":"pith_short_8","alias_value":"JKTQYFNE","created_at":"2026-05-18T12:30:25Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2016:JKTQYFNEHNRFYFZR7DN6FG67LW","target":"record","payload":{"canonical_record":{"source":{"id":"1612.04010","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2016-12-13T03:27:11Z","cross_cats_sorted":[],"title_canon_sha256":"a0c67bc14eae1774808058c00fefa9243d47863d3102a5f2f8b864275d6b5698","abstract_canon_sha256":"aad78584267f1d4319df830536f49ae22e8b7eb39da6ecbe19e4f6867a0fc6d2"},"schema_version":"1.0"},"canonical_sha256":"4aa70c15a43b625c1731f8dbe29bdf5d99a0bcb3212511e86210431de7c6aa5c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:28:32.372428Z","signature_b64":"bTW5aM+XIsgEOsWbR38rOb5WhvAv0GWLzrn8cv/TkNLDfkahq4Ee5v0JtgGl9nd20SorbDbCV9bI86jmrThjBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4aa70c15a43b625c1731f8dbe29bdf5d99a0bcb3212511e86210431de7c6aa5c","last_reissued_at":"2026-05-18T00:28:32.371744Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:28:32.371744Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1612.04010","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:28:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KMUoEcuEPjmj/dlT0D2OaBVZHNyr7xCB0kkviRVi7SEIrD3BnHFMCQocMeGWEz8DDh8yJgCH4i5GILhN5LBtAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T02:42:39.331520Z"},"content_sha256":"359176713d8193e547a3c51131853a4eb7ea641f572efed1122fb7b50caca77d","schema_version":"1.0","event_id":"sha256:359176713d8193e547a3c51131853a4eb7ea641f572efed1122fb7b50caca77d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2016:JKTQYFNEHNRFYFZR7DN6FG67LW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"An empirical analysis of the optimization of deep network loss surfaces","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daniel Jiwoong Im, Kristin Branson, Michael Tao","submitted_at":"2016-12-13T03:27:11Z","abstract_excerpt":"The success of deep neural networks hinges on our ability to accurately and efficiently optimize high-dimensional, non-convex functions. In this paper, we empirically investigate the loss functions of state-of-the-art networks, and how commonly-used stochastic gradient descent variants optimize these loss functions. To do this, we visualize the loss function by projecting them down to low-dimensional spaces chosen based on the convergence points of different optimization algorithms. Our observations suggest that optimization algorithms encounter and choose different descent directions at many "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1612.04010","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":""},"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:28:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bhf8D2DiUdrNHxecW037UI9Ck+1OZECOoGnF17Z5jynpgX0j35iKHePd5JGt3HIAbyrtSnSTJp8tRNVkFWNcBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T02:42:39.332190Z"},"content_sha256":"b51592d747e8b91f4e9728f280b4924a488288f8f1853dfaba2dfd9f1ac88842","schema_version":"1.0","event_id":"sha256:b51592d747e8b91f4e9728f280b4924a488288f8f1853dfaba2dfd9f1ac88842"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JKTQYFNEHNRFYFZR7DN6FG67LW/bundle.json","state_url":"https://pith.science/pith/JKTQYFNEHNRFYFZR7DN6FG67LW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JKTQYFNEHNRFYFZR7DN6FG67LW/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-26T02:42:39Z","links":{"resolver":"https://pith.science/pith/JKTQYFNEHNRFYFZR7DN6FG67LW","bundle":"https://pith.science/pith/JKTQYFNEHNRFYFZR7DN6FG67LW/bundle.json","state":"https://pith.science/pith/JKTQYFNEHNRFYFZR7DN6FG67LW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JKTQYFNEHNRFYFZR7DN6FG67LW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2016:JKTQYFNEHNRFYFZR7DN6FG67LW","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"aad78584267f1d4319df830536f49ae22e8b7eb39da6ecbe19e4f6867a0fc6d2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2016-12-13T03:27:11Z","title_canon_sha256":"a0c67bc14eae1774808058c00fefa9243d47863d3102a5f2f8b864275d6b5698"},"schema_version":"1.0","source":{"id":"1612.04010","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1612.04010","created_at":"2026-05-18T00:28:32Z"},{"alias_kind":"arxiv_version","alias_value":"1612.04010v4","created_at":"2026-05-18T00:28:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1612.04010","created_at":"2026-05-18T00:28:32Z"},{"alias_kind":"pith_short_12","alias_value":"JKTQYFNEHNRF","created_at":"2026-05-18T12:30:25Z"},{"alias_kind":"pith_short_16","alias_value":"JKTQYFNEHNRFYFZR","created_at":"2026-05-18T12:30:25Z"},{"alias_kind":"pith_short_8","alias_value":"JKTQYFNE","created_at":"2026-05-18T12:30:25Z"}],"graph_snapshots":[{"event_id":"sha256:b51592d747e8b91f4e9728f280b4924a488288f8f1853dfaba2dfd9f1ac88842","target":"graph","created_at":"2026-05-18T00:28:32Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"The success of deep neural networks hinges on our ability to accurately and efficiently optimize high-dimensional, non-convex functions. In this paper, we empirically investigate the loss functions of state-of-the-art networks, and how commonly-used stochastic gradient descent variants optimize these loss functions. To do this, we visualize the loss function by projecting them down to low-dimensional spaces chosen based on the convergence points of different optimization algorithms. Our observations suggest that optimization algorithms encounter and choose different descent directions at many ","authors_text":"Daniel Jiwoong Im, Kristin Branson, Michael Tao","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2016-12-13T03:27:11Z","title":"An empirical analysis of the optimization of deep network loss surfaces"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1612.04010","kind":"arxiv","version":4},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:359176713d8193e547a3c51131853a4eb7ea641f572efed1122fb7b50caca77d","target":"record","created_at":"2026-05-18T00:28:32Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"aad78584267f1d4319df830536f49ae22e8b7eb39da6ecbe19e4f6867a0fc6d2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2016-12-13T03:27:11Z","title_canon_sha256":"a0c67bc14eae1774808058c00fefa9243d47863d3102a5f2f8b864275d6b5698"},"schema_version":"1.0","source":{"id":"1612.04010","kind":"arxiv","version":4}},"canonical_sha256":"4aa70c15a43b625c1731f8dbe29bdf5d99a0bcb3212511e86210431de7c6aa5c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4aa70c15a43b625c1731f8dbe29bdf5d99a0bcb3212511e86210431de7c6aa5c","first_computed_at":"2026-05-18T00:28:32.371744Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:28:32.371744Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"bTW5aM+XIsgEOsWbR38rOb5WhvAv0GWLzrn8cv/TkNLDfkahq4Ee5v0JtgGl9nd20SorbDbCV9bI86jmrThjBQ==","signature_status":"signed_v1","signed_at":"2026-05-18T00:28:32.372428Z","signed_message":"canonical_sha256_bytes"},"source_id":"1612.04010","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:359176713d8193e547a3c51131853a4eb7ea641f572efed1122fb7b50caca77d","sha256:b51592d747e8b91f4e9728f280b4924a488288f8f1853dfaba2dfd9f1ac88842"],"state_sha256":"b92acdd314a0541eca8b56a23d6082aea55d0373966b0c800882a5db5ed1253d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oKK6OPSDUojZG2bGlLP+cINt548FBj0VIeRwz0BJ/U5s2OWB0TC5HPtECfvurX9ThHwxNxCTgJHkn2GyZeJ4DQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T02:42:39.335972Z","bundle_sha256":"1585a24ae0855ae5a1b089102541380a105e612a1fb26f289f2fe05f3e3400c2"}}