- Rating
- 1
- Confidence
- 4
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper studies the footprints of jailbreaking attacks by examining LLM early-decoding attention patterns. The proposed method constructs token-level anomaly scores from entropy-weighted attention profiles and robust within-prompt normalisation, enabling both prompt-level jailbreak detection and adversarial-span localisation without training an auxiliary classifier or updating the target model. Experiments on Llama-2 and Vicuna with GCG and AutoDAN attacks show high attack-detection recall and SpanHit rates, as well as reduced attack success rates when the detector is used as a gating mechanism.
Critical strengths
-
The motivation and method are clear, and the visualization is helpful.
-
The authors provide sufficient evidenc to support the effectiveness of their proposed method.
-
This paper is well-written, making it easy to follow.
Critical weaknesses
-
Insufficient evaluation of false positives and localisation accuracy. The use of only 17 benign controls is insufficient to reliably assess false positives, particularly for challenging safety-related but benign requests (e.g., harm-reduction or help-seeking queries, such as asking for assistance with drug cessation). In addition, the SpanHit metric is overly permissive, as any overlap with the annotated attack span counts as successful localisation, without evaluating precision or boundary accuracy.
-
Limited evaluation across harmful inputs and attack types. The evaluation focuses on GCG and AutoDAN, but the detector should also be systematically evaluated on plain harmful instructions (e.g., HarmBench or AdvBench) to clarify whether it detects harmfulness beyond adversarial modifications. Its robustness to qualitatively different attacks, such as PAIR and multilingual/translation-based jailbreaks, also remains unclear.
-
Limited cross-model transferability. Jailbreak behavior can be model-specific: the same prompt may successfully jailbreak one model but not another. It remains unclear whether attention footprints detected on one model (e.g., Llama) transfer to other models such as Qwen or API-only models. Cross-model evaluation on the same prompts would help establish whether the proposed signal is model-independent or model-specific.
-
Limited baselines and lack of frontier-model comparison. Strong contemporary safety models such as WildGuard, LlamaGuard, ShieldGemma, Qwen3Guard, or Shieldstral are not evaluated [1-5]. Moreover, zero-shot evaluation with strong frontier general-purpose LLMs (e.g., GPT-6 Astra) would also provide an important reference.
-
Generalisation to modern and larger models remains unclear. The experiments are restricted to Llama-2-7B/13B and Vicuna-13B, while contemporary models can differ substantially in architecture, alignment, and attention behaviour. Evaluation on stronger and larger models (e.g., recent Qwen 3.5 27B or MOE gpt-oss-20b) would be important, particularly because head selection is model-dependent. Moreover, the proposed approach requires white-box access to internal attention, limiting direct applicability to API-only models.
[1] Wildguard: Open one-stop moderation tools for safety risks, jailbreaks, and refusals of llms
[2] Llama guard: Llm-based input-output safeguard for human-ai conversations
[3] Shieldgemma: Generative ai content moderation based on gemma
[4] Qwen3guard technical report
[5] Shieldstral
Questions for authors
Refer to weaknesses.
We thank the reviewer for the detailed review. We ran the requested experiments; all numbers below use frozen detectors (no re-tuning on the new data).
W1: Benign controls and localization
We agree that 17 benign controls are insufficient. We added roughly 600–820 benign prompts per target: XSTest-safe, OR-Bench-Hard-1K (safety-related but benign), and Alpaca and Dolly prompts length-matched to GCG and AutoDAN.
The pooled benign false-positive rate is 45.0% on Llama-2-7B, 25.2% on Llama-2-13B and 2.0% on Vicuna-13B, against 0–1/17 in the paper.
W2: Plain harmful inputs and other attacks
The detector rarely fires on plain harmful requests: 0.5–3.8% on AdvBench and 0–4.7% on HarmBench direct requests across six models. So it does not detect harmfulness.
However, it also flags 15–35% of prompts carrying random, non-optimized suffixes. Detection of PAIR, TAP, persuasion and MultiJail attacks varies widely across models, from 0% to 99%. The signal therefore reflects unusual token spans rather than adversarial intent.
W3: Cross-model
On roughly 900–1,000 identical prompts evaluated on six models, detection decisions agree at chance level (Cohen’s κ between −0.10 and 0.12). The footprint is model-specific, so heads and thresholds must be selected per model, and white-box access is required.
W4: Baselines
On the same prompts (Llama-2-7B target), harmfulness guards cannot separate attacks from plain harmful requests:
W5: Modern models
On Llama-3-8B and Qwen2.5-7B/14B the results are model-dependent:
We will report these results together with the length limitation.