pith:PCRIFIFQ
Exponential Approximation Rates and Parameter Efficiency of Learnable Bernstein Activations
Learnable Bernstein polynomial activations achieve approximation error decaying as O(n^{-L}) with network depth and polynomial order, exponentially faster than ReLU networks while remaining fully differentiable.
arxiv:2602.04264 v2 · 2026-02-04 · cs.LG · cs.AI · cs.NA · math.NA
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{PCRIFIFQ3AKX2NASO54YF2G3MG}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
their approximation error decays with the network depth L and the polynomial order n with a rate of O(n^{-L}), exponentially faster than the polynomial rate of ReLU architectures while remaining fully differentiable.
The theoretical analysis assumes that the learnable Bernstein activations can be integrated into standard deep network architectures without introducing optimization difficulties that would negate the approximation benefits.
Learnable Bernstein activations achieve exponential approximation rates O(n^{-L}) and deliver up to 99.9% parameter reduction while matching or exceeding ReLU performance on large physics datasets.
Receipt and verification
| First computed | 2026-05-18T03:09:23.943437Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
78a282a0b0d8157d3412777982e8db619aa4a10971f47739c25c4edf2c67c76a
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/PCRIFIFQ3AKX2NASO54YF2G3MG \
| 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: 78a282a0b0d8157d3412777982e8db619aa4a10971f47739c25c4edf2c67c76a
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "daf496ce8d5fdc03fd988f3cbf5e60d5a7035e082ddce5807565ae7394c802f1",
"cross_cats_sorted": [
"cs.AI",
"cs.NA",
"math.NA"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"primary_cat": "cs.LG",
"submitted_at": "2026-02-04T06:55:55Z",
"title_canon_sha256": "9dca2bdd3ff3db2aeb9324011d1269d146505ca4273ea572b9e32bb75e3657ef"
},
"schema_version": "1.0",
"source": {
"id": "2602.04264",
"kind": "arxiv",
"version": 2
}
}