Pith Number

pith:6MVNH4QC

pith:2026:6MVNH4QCI2ZJKVB2TVSSMZH65B

not attested not anchored not stored refs resolved

DP-KFC: Data-Free Preconditioning for Privacy-Preserving Deep Learning

Albert Sund Aillet, Andrea Protani, Luigi Serio, Marc Molina Van den Bosch, Miguel Angel Gonzalez Ballester, Riccardo Taiello

DP-KFC constructs KFAC preconditioners from synthetic noise and frequency statistics alone.

arxiv:2605.13418 v1 · 2026-05-13 · cs.LG

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{6MVNH4QCI2ZJKVB2TVSSMZH65B}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

DP-KFC matches private-data preconditioners while public-data variants degrade by up to 4.8%, showing that curvature can be estimated without consuming privacy budget or introducing distribution shift.

C2weakest assumption

The Fisher Information Matrix decouples into architectural sensitivity recoverable via synthetic noise and input correlations approximable from modality-specific frequency statistics.

C3one line summary

DP-KFC approximates the Fisher Information Matrix for KFAC preconditioning via synthetic noise probes and modality frequency statistics, matching private-data performance without consuming privacy budget or introducing distribution shift.

References

16 extracted · 16 resolved · 0 Pith anchors

[1] Brendan and Mironov, Ilya and Talwar, Kunal and Zhang, Li , year= 1998 · doi:10.1145/2976749.2978318

[2] Ganesh, A., McMahan, B., and Thakurta, A 2025

[3] URL https://openreview.net/forum? id=j1zQGmQQOX1. Loshchilov, I. and Hutter, F. Decoupled weight decay reg- ularization. InInternational Conference on Learning Representations, 2019. URL https://openr 2019 · doi:10.1109/csf

[4] Thakkar, O., Andrew, G., and McMahan, H 1905 · doi:10.1609/aaai.v38i14.29451

[5] Linear Term (Descent): E[⟨∇L(θt),∆ t⟩] =E[⟨∇L(θ t),−η t(Pt¯gt +ξ t)⟩](10) =−η t⟨∇L(θt), PtE[¯gt]⟩ −η t⟨∇L(θt),E[ξ t]⟩(11) =−η t∇L(θt)⊤Pt∇L(θt)(SinceE[¯g t] =∇L,E[ξ t] = 0) (12) We use the spectral pro

Receipt and verification

First computed	2026-05-18T02:44:47.357751Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

f32ad3f20246b295543a9d652664fee85c31406b237ad59716321f3290e5c1f5

Aliases

arxiv: 2605.13418 · arxiv_version: 2605.13418v1 · doi: 10.48550/arxiv.2605.13418 · pith_short_12: 6MVNH4QCI2ZJ · pith_short_16: 6MVNH4QCI2ZJKVB2 · pith_short_8: 6MVNH4QC

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/6MVNH4QCI2ZJKVB2TVSSMZH65B \
  | 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: f32ad3f20246b295543a9d652664fee85c31406b237ad59716321f3290e5c1f5

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "185363e865ec7e407106d4e27b3728180945f1f6442be0abc4bea11b7f0c0dd3",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-13T12:14:00Z",
    "title_canon_sha256": "c8b162731120fd87715fee497add7f3baa41f2db6a10e54af1194844980893cb"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.13418",
    "kind": "arxiv",
    "version": 1
  }
}