Skip to content
OpenTrain AIFor AI Companies

HFEPX · Eval paper review

Phonetic Perturbations Reveal Tokenizer-Rooted Safety Gaps in LLMs

Darpan Aswal, Siddharth D Jaiswal

Published

May 20, 2025

Citations

0

Trust level

Low

Usefulness score

40/100 (Low)

Extraction confidence

45% (Low)

Derived from extracted protocol signals and abstract evidence.

Rater population

Not reported

Signals refreshed

Apr 7, 2026

Open arXiv Find implementation Shortlist (0) RSS feed

Tags

Coding Red Team

Should you rely on this paper?

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Use this as background context only. Do not make protocol decisions from this page alone.

Best use

Background context only

Use if you need

Background context only.

What to verify

Read the full paper before copying any benchmark, metric, or protocol choices.

Main weakness

The available metadata is too thin to trust this as a primary source.

Human feedback signal
Detected
From extracted signals
Evaluation signal
Weak or implicit
Validate from full paper
Usefulness for eval research
40/100
Adjacent candidate

Treat as adjacent context, not a core eval-method reference.

If you are doing eval pipeline work, start here

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

Abstract

Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics. We introduce CMP-RT (code-mixed phonetic perturbations for red-teaming), a novel diagnostic probe that pinpoints tokenization as the root cause of this vulnerability. A mechanistic analysis reveals that phonetic perturbations fragment safety-critical tokens into benign sub-words, suppressing their attribution scores while preserving prompt interpretability -- causing safety mechanisms to fail despite excellent input understanding. We demonstrate that this vulnerability evades standard defenses, persists across modalities and state-of-the-art (SOTA) models including Gemini-3-Pro, and scales through simple supervised fine-tuning (SFT). Furthermore, layer-wise probing shows perturbed and canonical input representations align up to a critical layer depth; enforcing output equivalence robustly recovers the lost representations, providing causal evidence for a structural gap between pre-training and alignment, and establishing tokenization as a critical, under-examined vulnerability in current safety pipelines.

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

partial

Red Team

Directly usable for protocol triage.

"Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics."

Evaluation Modes

missing

None explicit

Validate eval design from full paper text.

"Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics."

Quality Controls

missing

Not reported

No explicit QC controls found.

"Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics."

Benchmarks / Datasets

missing

Not extracted

No benchmark anchors detected.

"Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics."

Reported Metrics

missing

Not extracted

No metric anchors detected.

"Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics."

Benchmarks and datasets

No benchmark or dataset names were extracted from the available abstract.

Reported metrics

No metric terms were extracted from the available abstract.

Human feedback details
Uses human feedback
Yes
Feedback types
Red Team
Rater population
Not reported
Expertise required
Coding
Evaluation details
Evaluation modes
None
Agentic eval
None
Quality controls
Not reported
Evidence quality
Low
Use this page as
Background context only

Research brief

Metadata summary

Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics.

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

Key takeaways

  • Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics.
  • We introduce CMP-RT (code-mixed phonetic perturbations for red-teaming), a novel diagnostic probe that pinpoints tokenization as the root cause of this vulnerability.
  • A mechanistic analysis reveals that phonetic perturbations fragment safety-critical tokens into benign sub-words, suppressing their attribution scores while preserving prompt interpretability -- causing safety mechanisms to fail despite excellent input understanding.

Researcher actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Check the full text for explicit evaluation design choices (raters, protocol, and metrics).
  • 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

Contribution summary

  • Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics.
  • We introduce CMP-RT (code-mixed phonetic perturbations for red-teaming), a novel diagnostic probe that pinpoints tokenization as the root cause of this vulnerability.
  • We demonstrate that this vulnerability evades standard defenses, persists across modalities and state-of-the-art (SOTA) models including Gemini-3-Pro, and scales through simple supervised fine-tuning (SFT).

Why it matters for eval

  • Safety-aligned LLMs remain vulnerable to digital phenomena like textese that introduce non-canonical perturbations to words but preserve the phonetics.
  • A mechanistic analysis reveals that phonetic perturbations fragment safety-critical tokens into benign sub-words, suppressing their attribution scores while preserving prompt interpretability -- causing safety mechanisms to fail despite…

Researcher checklist

  • Human feedback protocol is explicit

    Detected: Red Team

  • Evaluation mode is explicit

    No clear evaluation mode extracted.

  • Quality control reporting appears

    No calibration/adjudication/IAA control explicitly detected.

  • Benchmark or dataset anchors are present

    No benchmark/dataset anchor extracted from abstract.

  • Metric reporting is present

    No metric terms extracted.