CommunityJev Practices & Patterns
HydroJEV: A one-second, training-free screen for cyber-attack and fault attribution in water distribution networks (arXiv 2610.02048)
A pre-registered evaluation of Jev as a one-second, training-free screen for SCADA alarm triage in water distribution networks. On a four-class EPANET C-Town benchmark it matched a hand-written rule tree, and a gated cascade spared an LLM reviewer 35-38% of windows.
- Format
- Article
- Published by
- Community
- Last verified
- Oct 2026
Highlights
- Triage assigns each alarm one of four causes: cyberattack, physical fault, normal transient, or faulty sensor.
- Jev matched the hand-written rule tree at macro-F1 0.62-0.64 against 0.56-0.61 in distribution with label-free prior correction.
- Beat the supervised classifier by 0.36-0.42 on event subtypes absent from its labels, in all four sealed pre-registered rounds.
- Decided 20-40 times faster than frontier LLMs and outperformed the classifier below about four labelled events per class.
- A rule-tree-gated cascade spared an LLM reviewer 35-38% of windows on fresh sealed sets without macro-F1 loss.
Watch out
arXiv preprint posted October 1, 2026 (cs.AI), 41 pages, 19 figures. Results cover one simulated EPANET network (C-Town) plus two transfer networks, all in the authors' sealed rounds; the cascade stayed within a non-inferiority margin, but the abstract reports no real utility SCADA deployments.
Questions about this entry
Short answers drawn from the linked project or page.
It screens SCADA alarms into four causes: cyberattack, physical fault, normal transient, or a faulty sensor, so operators can see what needs deliberate review.
The authors built a four-class cause-attribution benchmark on the C-Town network in EPANET and later transferred the unchanged gated cascade to two further networks.
With only a label-free prior correction, Jev matched the rule tree at macro-F1 0.62-0.64 against 0.56-0.61 in distribution, in all four rounds.
Jev exceeded the supervised classifier by 0.36-0.42 macro-F1 on event subtypes absent from its labels, in all four sealed, pre-registered rounds.
Accepting only benign Jev verdicts confirmed by the rule tree spared an LLM reviewer 35-38% of windows on fresh sealed sets without losing macro-F1.
Jev returns class probabilities in about one second and decided 20-40 times faster than the seven cloud LLMs it faced on identical evidence in the sealed rounds.
Jev outperformed the supervised classifier whenever fewer than about four labelled events per class were available, the label-scarce setting utilities face.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
YouTube: Kubernetes admission control with a decision model, fail-closed.
Hacker News: SafeDep Gryph plus Jev checks for compromised coding-agent behavior.
2 pointsDetecting Compromised AI Coding Agents with Jev and GryphRead the thread on Hacker NewsHacker News: Show HN: Agent Chaperone screens tool calls and results with Jev before the agent sees them.
sepehrsafari5 pointsShow HN: Agent Chaperone – Screen AI agent tool calls and results with JevRead the thread on Hacker NewsX: Secret detection with Jev, tested
Related terms
Glossary definitions related to this entry.
More Jev resources like this
A position and case-study paper on using System One decision models like Jev and the open-source Laya as decision layers in LLM-driven penetration-testing harnesses. It names four decision points, runs an exploratory 13-vulnerability NeuroSploit comparison, and proposes Rave, an unevaluated model.
Jev Practices & PatternsArticle#community#research#arxiv
Community
Documents a failure mode in letter-bound remote sensing classification: 97.71% OA on UCM in vocabulary but 7.35%–96.40% cross-dataset accuracy and inverted confidence (AUROC 0.38). RSAJEV replaces the letter readout with training-free prototype retrieval, reaching 89.43% on the UCM cross-dataset gate.
Jev Practices & PatternsArticle#community#research#arxiv
Community
3GitHub stars
Blind security benchmarks for Jev on public corpora: 662 prompt-injection messages and 200 matched vulnerable-code pairs, with raw per-sample results and a TUI dashboard.
Jev Practices & Patterns#community#go#benchmarks
Community
Gaurav-Gosain/jev-sec-benchGo
Gaurav-Gosain/jev-sec-benchGo4GitHub stars
An analysis of ordinal-scale bias in JEV 1.13 and three open KEV models: across 36 ordinal datasets, final decisions use only 67-76% of the effective gold support versus 87-102% on four nominal tasks. BA-LoRA post-training raises utilization from about 47% to 86%.
Jev Practices & PatternsArticle#community#research#arxiv
Community
Glax147/jev_ordinal_scale_biaPython
Glax147/jev_ordinal_scale_biaPython3GitHub stars
PACT turns contrastive training pairs into four loss terms for single-token typed-decision models. On a frozen 324-item holdout it matches the published recipe's accuracy (84.6% versus 85.2%) while cutting answer flips under relabelling from 13.8% to 9.8%.
Jev Practices & PatternsArticle#community#research#arxiv
An evaluation of Jev and two open-weight typed decision models when option names and definitions conflict. With yes/no names, decision-flip rates are up to 70.4 percentage points higher than the 0/1 control, and mean AUC drops from 93.8% to a below-chance 23.2%.
Jev Practices & PatternsArticle#community#research#arxiv
Community
From the Jev guides
Original write-ups on what this entry is for.
More than eighty preprints now evaluate, attack, and reproduce the typed-decision class. The broadest board, the coherence gap, the rejection bottleneck, and the attacks that still work.
Jev Practices & Patterns#research#evaluation#calibration
Read article
TypeSafe publishes evals, but the numbers that matter are the ones measured on your task. A tour of the independent benchmarks, field reports, and the recurring finding that question design dominates.
Jev Practices & Patterns#benchmarks#evaluation#calibration
Read article
A Choice can only pick from its options. How rejection options, candidate coverage, and escalation bands keep a narrow taxonomy honest under inputs it never anticipated.
Jev Practices & Patterns#rejection#confidence#patterns
Read article
From the Jev community
Posts from builders shipping with Jev right now.