pith:HPZ2TYLU
Robust and Fast Training via Per-Sample Clipping
Per-sample gradient clipping in SGD achieves optimal convergence rates for non-convex problems under heavy-tailed noise.
arxiv:2605.02701 v2 · 2026-05-04 · math.OC · cs.LG · stat.ML
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{HPZ2TYLUIIQHXHRG4FXIOJQXM6}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
We show that the resulting method, per-sample clipped SGD (PS-Clip-SGD), achieves optimal in-expectation convergence rates for non-convex optimization problems under heavy-tailed gradient noise. Moreover, we establish high-probability convergence guarantees that match the in-expectation rates up to polylogarithmic factors in the failure probability.
The analysis requires that gradient noise follows a heavy-tailed distribution with specific moment bounds; if real gradients during training do not satisfy these tail conditions, the claimed optimal rates may not apply.
Per-sample clipped SGD achieves optimal in-expectation and high-probability convergence rates for non-convex optimization under heavy-tailed gradient noise while outperforming standard SGD and batch clipping on CIFAR-100.
Formal links
Receipt and verification
| First computed | 2026-06-24T01:15:03.316448Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
3bf3a9e17442207b9e26e16e87261767bc066827e81d9a2eca60b588c355868c
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/HPZ2TYLUIIQHXHRG4FXIOJQXM6 \
| 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: 3bf3a9e17442207b9e26e16e87261767bc066827e81d9a2eca60b588c355868c
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "81483593c864179e7b4f2c20037d60301afd256c002c88ce94643f81ab0195c7",
"cross_cats_sorted": [
"cs.LG",
"stat.ML"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "math.OC",
"submitted_at": "2026-05-04T15:11:36Z",
"title_canon_sha256": "8d9580df019da6feb4704f296d90c73bb3b3f3d4ba7f32d36c8da0007da41cfc"
},
"schema_version": "1.0",
"source": {
"id": "2605.02701",
"kind": "arxiv",
"version": 2
}
}