pith:5RZVF2NG
Tighter Information-Theoretic Generalization Bounds via a Novel Class of Change of Measure Inequalities
Novel change of measure inequalities derived from the data processing inequality provide tighter bounds for generalization and privacy.
arxiv:2602.07999 v4 · 2026-02-08 · cs.IT · cs.LG · math.IT
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{5RZVF2NGPNHX5PEXSOIKKNY3GN}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
We propose novel change of measure inequalities via a unified framework based on the data processing inequality, which is surprisingly elementary yet powerful enough to yield novel, tighter inequalities.
The data-processing inequality can be applied directly to the chosen family of information measures without additional regularity conditions that would invalidate the claimed tightness in the generalization and privacy settings.
A unified data-processing framework produces tighter change-of-measure inequalities that improve information-theoretic generalization bounds across learning theory and privacy.
Formal links
Receipt and verification
| First computed | 2026-05-28T01:04:36.582954Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
ec7352e9a67b4f7ebc979390a5371b33541edf8c52b92b3552a812075efc2946
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/5RZVF2NGPNHX5PEXSOIKKNY3GN \
| 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: ec7352e9a67b4f7ebc979390a5371b33541edf8c52b92b3552a812075efc2946
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "edbd8325cb6e507ceb69775a9f708dbde2e034c9e6958024415aa4d5c9f3594b",
"cross_cats_sorted": [
"cs.LG",
"math.IT"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"primary_cat": "cs.IT",
"submitted_at": "2026-02-08T14:53:14Z",
"title_canon_sha256": "9c4d90faa71b5c388a348a66dcc2365b4cb313fccc88f3ff0d226dd8000eea20"
},
"schema_version": "1.0",
"source": {
"id": "2602.07999",
"kind": "arxiv",
"version": 4
}
}