pith:ZS2SKAOO
Zeroth-Order Optimization at the Edge of Stability
Zeroth-order methods remain mean-square stable only when their step size satisfies a bound that depends on the full Hessian spectrum rather than its largest eigenvalue alone.
arxiv:2604.14669 v2 · 2026-04-16 · cs.LG · math.DS · math.OC · stat.ML
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{ZS2SKAOOPR7PRYBOJLEYQNVX2I}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
we provide an explicit step size condition that exactly captures the (mean-square) linear stability of a family of ZO methods based on the standard two-point estimator... mean-square stability of ZO methods depends on the entire Hessian spectrum.
The linear stability analysis assumes the loss can be locally approximated by a quadratic form whose Hessian is constant over the relevant trajectory, and that the two-point estimator is used without additional smoothing or variance reduction.
Zeroth-order methods achieve mean-square stability when the step size satisfies a condition involving the entire Hessian spectrum, with full-batch ZO optimizers operating at the edge of stability and large steps regularizing the Hessian trace.
Receipt and verification
| First computed | 2026-07-02T01:17:31.259570Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
ccb52501ce7c7ef8e02e4ac98836b7d209df2144765ef2366c12dc9fb3ae21d1
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/ZS2SKAOOPR7PRYBOJLEYQNVX2I \
| jq -c '.canonical_record' \
| python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: ccb52501ce7c7ef8e02e4ac98836b7d209df2144765ef2366c12dc9fb3ae21d1
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "1e3b5b5f0d85cc6d5838d159f8430a92acd19e59916cfe0e1b08dc4e9ac8fc9f",
"cross_cats_sorted": [
"math.DS",
"math.OC",
"stat.ML"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.LG",
"submitted_at": "2026-04-16T06:23:18Z",
"title_canon_sha256": "579a22f9f3d9dd0870be92dac9c81190f376974fd9f771bf37ac8126983d8a4f"
},
"schema_version": "1.0",
"source": {
"id": "2604.14669",
"kind": "arxiv",
"version": 2
}
}