Pith Number

pith:I6F6HXEC

pith:2014:I6F6HXECCAG7OT4NJRVLVOUIFF

not attested not anchored not stored refs pending

Explaining and Harnessing Adversarial Examples

Christian Szegedy, Ian J. Goodfellow, Jonathon Shlens

Neural networks are vulnerable to adversarial examples mainly because they behave linearly in their inputs.

arxiv:1412.6572 v3 · 2014-12-20 · stat.ML · 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{I6F6HXECCAG7OT4NJRVLVOUIFF}

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

We argue instead that the primary cause of neural networks' vulnerability to adversarial perturbation is their linear nature.

C2weakest assumption

That the network's output is sufficiently linear in the input for small perturbations so that a first-order approximation accurately predicts the effect of the perturbation.

C3one line summary

Neural networks' susceptibility to adversarial examples stems from their linear nature, enabling a fast sign-of-gradient method for generating perturbations that also supports adversarial training to boost robustness.

Formal links

2 machine-checked theorem links

Cited by

271 papers in Pith

LiST: Lipschitz Scaling Training for Robust and Calibrated Neural Networks

Mechanistic Interpretability of LLM Jailbreaks via Internal Attribution Graphs

Efficient Safety Alignment of Language Models via Latent Personality Traits

Tubular Neighbourhoods of Pfaffian Sets and Applications to Neural Networks

Gimitest: A Comprehensive Tool for Testing Reinforcement Learning Policies

Receipt and verification

First computed	2026-07-04T19:26:30.839514Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

478be3dc82100df74f8d4c6ababa88295bc68e77d0cc3da2ef932f58489d3b99

Aliases

arxiv: 1412.6572 · arxiv_version: 1412.6572v3 · doi: 10.48550/arxiv.1412.6572 · pith_short_12: I6F6HXECCAG7 · pith_short_16: I6F6HXECCAG7OT4N · pith_short_8: I6F6HXEC

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/I6F6HXECCAG7OT4NJRVLVOUIFF \
  | 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: 478be3dc82100df74f8d4c6ababa88295bc68e77d0cc3da2ef932f58489d3b99

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "d1cab7a040ad5c17c6bab8f4f7bd8444fa99bd96046bcd23b5f2ffefd428de73",
    "cross_cats_sorted": [
      "cs.LG"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "stat.ML",
    "submitted_at": "2014-12-20T01:17:12Z",
    "title_canon_sha256": "6d7dd56a8a46845d10b69a881dab23f4ae149762ccd7d3c067c4e6a44f354796"
  },
  "schema_version": "1.0",
  "source": {
    "id": "1412.6572",
    "kind": "arxiv",
    "version": 3
  }
}