pith:5BLQ2JLI
Sharp description of local minima in the loss landscape of high-dimensional two-layer ReLU neural networks
Local minima in two-layer ReLU networks admit an exact low-dimensional representation via summary statistics.
arxiv:2604.09412 v2 · 2026-04-10 · stat.ML · cond-mat.dis-nn · cs.LG
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{5BLQ2JLIEI5VQQA7JOK76IH3TA}
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 local minima admit an exact low-dimensional representation in terms of summary statistics, yielding a sharp and interpretable characterisation of the landscape. We further establish a direct link with one-pass SGD: local minima correspond to attractive fixed points of the dynamics in summary statistics space.
The analysis assumes a realizable teacher-student setting with Gaussian covariates and the specific network form sum ReLU(w_k^T x), which may not hold for more general data distributions or deeper networks.
Local minima of two-layer ReLU networks have an exact representation in terms of a few summary statistics that correspond to attractive fixed points of one-pass SGD, with overparameterization connecting them via flat directions.
Receipt and verification
| First computed | 2026-06-01T01:03:52.799059Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
e8570d2568223b58401f4b95ff20fb9834af54541f32d74dd9429d3c49586448
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/5BLQ2JLIEI5VQQA7JOK76IH3TA \
| 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: e8570d2568223b58401f4b95ff20fb9834af54541f32d74dd9429d3c49586448
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "2b457ee33e7d83e54e093d13c42311766f8d2c24a201b3cd3dc40cb9add9b499",
"cross_cats_sorted": [
"cond-mat.dis-nn",
"cs.LG"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"primary_cat": "stat.ML",
"submitted_at": "2026-04-10T15:26:00Z",
"title_canon_sha256": "2758a56d67f00d43bd6fb4300fc3a3330478542021269a49b3f30e34303d05b3"
},
"schema_version": "1.0",
"source": {
"id": "2604.09412",
"kind": "arxiv",
"version": 2
}
}