pith:HWR2B3TF
A projection-based framework for gradient-free and parallel learning
Neural network training can be recast as projecting parameters onto local constraints from each operation instead of using gradients.
arxiv:2506.05878 v3 · 2025-06-06 · cs.LG
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{HWR2B3TF4DD2CAN2KQSAKH373P}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
We reformulate training as a large-scale feasibility problem: finding network parameters and states that satisfy local constraints derived from its elementary operations. Training then involves projecting onto these constraints, a local operation that can be parallelized across the network.
That iterative projection algorithms applied to the composed local constraints will converge to useful network parameters that achieve competitive performance on standard benchmarks, as claimed in the demonstration of PJAX on MLPs, CNNs, and RNNs.
Neural network training is recast as a large-scale feasibility problem solved by composing and iterating projection operators, implemented in a new JAX framework called PJAX.
Formal links
Receipt and verification
| First computed | 2026-05-20T00:00:20.868308Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
3da3a0ee65e0c7a101ba5424051f7fdbe904b1b0d42a35b917f47fac16b0f9a5
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/HWR2B3TF4DD2CAN2KQSAKH373P \
| 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: 3da3a0ee65e0c7a101ba5424051f7fdbe904b1b0d42a35b917f47fac16b0f9a5
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "7cc6083d4c3c2155d7181744a6353df264095aab098c64ca274e5cf630e9006a",
"cross_cats_sorted": [],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.LG",
"submitted_at": "2025-06-06T08:44:56Z",
"title_canon_sha256": "5c3f8bbbf5f04f514e074c4999dcf318e101aaed9961118f1f93979e4f86a626"
},
"schema_version": "1.0",
"source": {
"id": "2506.05878",
"kind": "arxiv",
"version": 3
}
}