Skip to content
OpenTrain AIFor AI Companies

HFEPX · Eval paper review

Do Encoders Suffice? A Systematic Comparison of Encoder and Decoder Safety Judges for LLM Adversarial Evaluation

Han Jeon, Shiv Medler, Joseph Voyles, Matt Wood

Published

Jun 24, 2026

Citations

0

Trust level

High

Usefulness score

77/100 (High)

Extraction confidence

80% (High)

Derived from extracted protocol signals and abstract evidence.

Rater population

Not reported

Signals refreshed

Jun 24, 2026

Open arXiv Find implementation Shortlist (0) RSS feed

Tags

Automatic Metrics General Llm As Judge Red Team

Should you rely on this paper?

This paper has strong direct human-feedback and evaluation protocol signal and is suitable as a primary eval pipeline reference.

Use this as a practical starting point for protocol research, then validate against the original paper.

Best use

Primary benchmark and eval reference

Use if you need

A benchmark-and-metrics comparison anchor.

What to verify

Validate the evaluation procedure and quality controls in the full paper before operational use.

Main weakness

No major weakness surfaced.

Human feedback signal
Detected
From extracted signals
Evaluation signal
Detected
Eval setup described
Usefulness for eval research
77/100
High-confidence candidate

Use this as a primary source when designing or comparing eval protocols.

If you are doing eval pipeline work, start here

Human Eval Papers LLM-as-Judge Papers Pairwise Preference Papers Tool-Use Evaluation Search

Abstract

With the widespread adoption of large language models (LLMs) in chatbots and everyday applications, companies increasingly need guardrails that are effective while remaining low-cost and low-latency. Safety evaluation of LLM outputs has generally relied on LLM-based judges, which can be effective but are often slow and expensive to deploy at scale. In this paper, we evaluate whether fine-tuned modern encoder classifiers from the ModernBERT family, including ModernBERT and Ettin, can reliably identify harmful LLM outputs in user-model conversations without substantial performance loss relative to LLM-based judges. We benchmark these encoder classifiers against rule-based prefix matching, fine-tuned LLM classifiers, and LLM judges using a range of judge-prompting strategies across open-source adversarial datasets. The LLM judges include evaluation methodologies from StrongReject, ShieldGemma, JailbreakBench, AILuminate, SorryBench, and a Claude-as-a-judge setup, as well as fine-tuned safety classifiers such as LlamaGuard 3 and LlamaGuard 4. The encoder classifiers are fine-tuned on judge-labeled data using a majority-voting label strategy and are then evaluated on a gold-standard holdout dataset to assess their performance relative to LLM judges. We report absolute performance using F1 score, false negative rate, and precision-recall metrics. We also break down results by attack technique, including single-turn prompting, decomposition, escalation, and context manipulation, to identify where encoder classifiers align with or diverge from LLM-based judges. Our findings provide guidance on when encoder classifiers can serve as cost- and latency-efficient alternatives to LLM-based safety evaluation.

What we could verify

These are the protocol signals we could actually recover from the available paper metadata. Use them to decide whether this paper is worth deeper reading.

Human Feedback Types

strong

Red Team

Directly usable for protocol triage.

"With the widespread adoption of large language models (LLMs) in chatbots and everyday applications, companies increasingly need guardrails that are effective while remaining low-cost and low-latency."

Evaluation Modes

strong

Llm As Judge, Automatic Metrics

Includes extracted eval setup.

"With the widespread adoption of large language models (LLMs) in chatbots and everyday applications, companies increasingly need guardrails that are effective while remaining low-cost and low-latency."

Quality Controls

missing

Not reported

No explicit QC controls found.

"With the widespread adoption of large language models (LLMs) in chatbots and everyday applications, companies increasingly need guardrails that are effective while remaining low-cost and low-latency."

Benchmarks / Datasets

strong

Jailbreakbench, Sorrybench

Useful for quick benchmark comparison.

"The LLM judges include evaluation methodologies from StrongReject, ShieldGemma, JailbreakBench, AILuminate, SorryBench, and a Claude-as-a-judge setup, as well as fine-tuned safety classifiers such as LlamaGuard 3 and LlamaGuard 4."

Reported Metrics

strong

F1, Precision, Recall

Useful for evaluation criteria comparison.

"We report absolute performance using F1 score, false negative rate, and precision-recall metrics."

Benchmarks and datasets

JailbreakbenchSorrybench

Reported metrics

f1precisionrecall
Human feedback details
Uses human feedback
Yes
Feedback types
Red Team
Rater population
Not reported
Expertise required
General
Evaluation details
Evaluation modes
Llm As Judge, Automatic Metrics
Agentic eval
None
Quality controls
Not reported
Evidence quality
High
Use this page as
Primary benchmark and eval reference

Research brief

Metadata summary

With the widespread adoption of large language models (LLMs) in chatbots and everyday applications, companies increasingly need guardrails that are effective while remaining low-cost and low-latency.

Based on abstract + metadata only. Check the source paper before making high-confidence protocol decisions.

Key takeaways

  • With the widespread adoption of large language models (LLMs) in chatbots and everyday applications, companies increasingly need guardrails that are effective while remaining low-cost and low-latency.
  • Safety evaluation of LLM outputs has generally relied on LLM-based judges, which can be effective but are often slow and expensive to deploy at scale.
  • In this paper, we evaluate whether fine-tuned modern encoder classifiers from the ModernBERT family, including ModernBERT and Ettin, can reliably identify harmful LLM outputs in user-model conversations without substantial performance loss relative to LLM-based judges.

Researcher actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Validate inferred eval signals (Automatic metrics) against the full paper.
  • Use related-paper links to find stronger protocol-specific references.

Caveats

  • Generated from abstract + metadata only; no PDF parsing.
  • Signals below are heuristic and may miss details reported outside the abstract.

Recommended queries

F1

Contribution summary

  • Safety evaluation of LLM outputs has generally relied on LLM-based judges, which can be effective but are often slow and expensive to deploy at scale.
  • In this paper, we evaluate whether fine-tuned modern encoder classifiers from the ModernBERT family, including ModernBERT and Ettin, can reliably identify harmful LLM outputs in user-model conversations without substantial performance loss…
  • We benchmark these encoder classifiers against rule-based prefix matching, fine-tuned LLM classifiers, and LLM judges using a range of judge-prompting strategies across open-source adversarial datasets.

Why it matters for eval

  • Safety evaluation of LLM outputs has generally relied on LLM-based judges, which can be effective but are often slow and expensive to deploy at scale.
  • In this paper, we evaluate whether fine-tuned modern encoder classifiers from the ModernBERT family, including ModernBERT and Ettin, can reliably identify harmful LLM outputs in user-model conversations without substantial performance loss…

Researcher checklist

  • Human feedback protocol is explicit

    Detected: Red Team

  • Evaluation mode is explicit

    Detected: Llm As Judge, Automatic Metrics

  • Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Benchmark or dataset anchors are present

    Detected: Jailbreakbench, Sorrybench

  • Metric reporting is present

    Detected: f1, precision, recall