pith:IFMVKYUV
Beyond Pattern Matching: Seven Cross-Domain Techniques for Prompt Injection Detection
Seven techniques borrowed from bioinformatics, linguistics, and other fields detect prompt injections more effectively than regex or classifiers.
arxiv:2604.18248 v2 · 2026-04-20 · cs.CR · cs.CL
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{IFMVKYUVDSBEUSCO6GW2YVUWJD}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
The local-alignment detector lifts F1 on deepset from 0.033 to 0.378 with zero additional false positives. The stylometric detector adds 11.1 percentage points of F1 on an indirect-injection benchmark.
That mechanisms proven in their original domains (sequence alignment, fatigue analysis, etc.) will transfer to LLM prompt injection without being bypassed by adaptive adversaries or introducing new failure modes not captured in the six evaluation datasets.
Seven cross-domain techniques for prompt injection detection are proposed; three implemented versions raise F1 scores on multiple benchmarks while releasing all code and data.
Receipt and verification
| First computed | 2026-05-20T00:05:44.845693Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
41595562951c824a484ef1adac569648fc63092c2099d1355cd59c24408cc4f1
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/IFMVKYUVDSBEUSCO6GW2YVUWJD \
| 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: 41595562951c824a484ef1adac569648fc63092c2099d1355cd59c24408cc4f1
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "65072f82d2b9aac8b5d15547c412646aa1df0076e9143dc99c9a04008da479fd",
"cross_cats_sorted": [
"cs.CL"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.CR",
"submitted_at": "2026-04-20T13:27:05Z",
"title_canon_sha256": "22ee573876f9c2fc734740f96a98698e5de844d0d13b190286e8cbf7950ed3d5"
},
"schema_version": "1.0",
"source": {
"id": "2604.18248",
"kind": "arxiv",
"version": 2
}
}