Searcharxiv⌕ Search

arXiv subjects

Chawin Sitawarin

Publications and source records attributed to Chawin Sitawarin.

At least 19 recordsLinked to original sources
arXiv2026

Rapid Poison: Practical Poisoning Attacks Against the Rapid Response Framework

The Rapid Response (RR) framework, deployed in production systems, including Anthropic's ASL-3 safeguards, continuously improves jailbreak-detection classifiers. When new jailbreaks emerge that bypass these classifiers, Rapid Response generates synthetic variants for training, helping the model generalize from the new attacks and quickly adapt. We reveal that prompt injection can infiltrate this pipeline to deliver poisoned samples into the classifier's training set, enabling two attack objectives: (I) targeted poisoning attacks that create false positives on harmless samples by categorizing them as a jailbreak, with a specific desired feature (e.g., certain formatting, subject, or keyword), (II) concept-based backdoor attacks that induce false negatives on jailbreak inputs, generalizing even to jailbreaks from attack strategies the defender explicitly trained against, when the backdoor trigger is present. Importantly, our threat model restricts adversaries to modifying only jailbreak samples (not benign data or labels), a constraint unexplored by prior work that makes the second objective particularly challenging. We address this with Omission Attack, which exploits a new phenomenon: when training on concept-absent unsafe samples, the classifier misassociates that concept's presence with the safe label. Both attacks cause substantial and in some cases near-complete label flipping at only a 1% poisoning rate, achieving up to 100% false positive rates and up to 96% false negative rates.

cs.LG↗
arXiv2026

Soft Instruction De-escalation Defense

Large Language Models (LLMs) are increasingly deployed in agentic systems that interact with an external environment; this makes them susceptible to prompt injections when dealing with untrusted data. To overcome this limitation, we propose SIC (Soft Instruction Control)-a simple yet effective iterative prompt sanitization loop designed for tool-augmented LLM agents. Our method repeatedly inspects incoming data for instructions that could compromise agent behavior. If such content is found, the malicious content is rewritten, masked, or removed, and the result is re-evaluated. The process continues until the input is clean or a maximum iteration limit is reached; if imperative instruction-like content remains, the agent halts to ensure security. By allowing multiple passes, our approach acknowledges that individual rewrites may fail but enables the system to catch and correct missed injections in later steps. Although immediately useful, worst-case analysis shows that SIC is not infallible; strong adversary can still get a 15% ASR by embedding non-imperative workflows. This nonetheless raises the bar.

cs.CR↗
arXiv2026

Positional Embedding-Aware Activations

We present a neural network architecture designed to naturally learn a positional embedding and overcome the spectral bias towards lower frequencies faced by conventional activation functions. Our proposed architecture, SPDER, is a simple MLP that uses an activation function composed of a sinusoidal multiplied by a sublinear function, called the damping function. The sinusoidal enables the network to automatically learn the positional embedding of an input coordinate while the damping passes on the actual coordinate value by preventing it from being projected down to within a finite range of values. Our results indicate that SPDERs speed up training by 10x and converge to losses 1,500-50,000x lower than that of the state-of-the-art for image representation. SPDER is also state-of-the-art in audio representation. The superior representation capability allows SPDER to also excel on multiple downstream tasks such as image super-resolution and video frame interpolation. We provide intuition as to why SPDER significantly improves fitting compared to that of other INR methods while requiring no hyperparameter tuning or preprocessing.

cs.CV↗
arXiv2025

Extracting alignment data in open models

In this work, we show that it is possible to extract significant amounts of alignment training data from a post-trained model -- useful to steer the model to improve certain capabilities such as long-context reasoning, safety, instruction following, and maths. While the majority of related work on memorisation has focused on measuring success of training data extraction through string matching, we argue that embedding models are better suited for our specific goals. Distances measured through a high quality embedding model can identify semantic similarities between strings that a different metric such as edit distance will struggle to capture. In fact, in our investigation, approximate string matching would have severely undercounted (by a conservative estimate of $10\times$) the amount of data that can be extracted due to trivial artifacts that deflate the metric. Interestingly, we find that models readily regurgitate training data that was used in post-training phases such as SFT or RL. We show that this data can be then used to train a base model, recovering a meaningful amount of the original performance. We believe our work exposes a possibly overlooked risk towards extracting alignment data. Finally, our work opens up an interesting discussion on the downstream effects of distillation practices: since models seem to be regurgitating aspects of their training set, distillation can therefore be thought of as indirectly training on the model's original dataset.

cs.AI↗
arXiv2025

The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections

How should we evaluate the robustness of language model defenses? Current defenses against jailbreaks and prompt injections (which aim to prevent an attacker from eliciting harmful knowledge or remotely triggering malicious actions, respectively) are typically evaluated either against a static set of harmful attack strings, or against computationally weak optimization methods that were not designed with the defense in mind. We argue that this evaluation process is flawed. Instead, we should evaluate defenses against adaptive attackers who explicitly modify their attack strategy to counter a defense's design while spending considerable resources to optimize their objective. By systematically tuning and scaling general optimization techniques-gradient descent, reinforcement learning, random search, and human-guided exploration-we bypass 12 recent defenses (based on a diverse set of techniques) with attack success rate above 90% for most; importantly, the majority of defenses originally reported near-zero attack success rates. We believe that future defense work must consider stronger attacks, such as the ones we describe, in order to make reliable and convincing claims of robustness.

cs.LG↗
arXiv2025

Defending Against Prompt Injection With a Few DefensiveTokens

When large language model (LLM) systems interact with external data to perform complex tasks, a new attack, namely prompt injection, becomes a significant threat. By injecting instructions into the data accessed by the system, the attacker is able to override the initial user task with an arbitrary task directed by the attacker. To secure the system, test-time defenses, e.g., defensive prompting, have been proposed for system developers to attain security only when needed in a flexible manner. However, they are much less effective than training-time defenses that change the model parameters. Motivated by this, we propose DefensiveToken, a test-time defense with prompt injection robustness comparable to training-time alternatives. DefensiveTokens are newly inserted as special tokens, whose embeddings are optimized for security. In security-sensitive cases, system developers can append a few DefensiveTokens before the LLM input to achieve security with a minimal utility drop. In scenarios where security is less of a concern, developers can simply skip DefensiveTokens; the LLM system remains the same as there is no defense, generating high-quality responses. Thus, DefensiveTokens, if released alongside the model, allow a flexible switch between the state-of-the-art (SOTA) utility and almost-SOTA security at test time. The code is available at https://github.com/Sizhe-Chen/DefensiveToken.

cs.CR↗
arXiv2025

Does More Inference-Time Compute Really Help Robustness?

Recently, Zaremba et al. demonstrated that increasing inference-time computation improves robustness in large proprietary reasoning LLMs. In this paper, we first show that smaller-scale, open-source models (e.g., DeepSeek R1, Qwen3, Phi-reasoning) can also benefit from inference-time scaling using a simple budget forcing strategy. More importantly, we reveal and critically examine an implicit assumption in prior work: intermediate reasoning steps are hidden from adversaries. By relaxing this assumption, we identify an important security risk, intuitively motivated and empirically verified as an inverse scaling law: if intermediate reasoning steps become explicitly accessible, increased inference-time computation consistently reduces model robustness. Finally, we discuss practical scenarios where models with hidden reasoning chains are still vulnerable to attacks, such as models with tool-integrated reasoning and advanced reasoning extraction attacks. Our findings collectively demonstrate that the robustness benefits of inference-time scaling depend heavily on the adversarial setting and deployment context. We urge practitioners to carefully weigh these subtle trade-offs before applying inference-time scaling in security-sensitive, real-world applications.

cs.AI↗
arXiv2025

How much do language models memorize?

We propose a new method for estimating how much a model knows about a datapoint and use it to measure the capacity of modern language models. Prior studies of language model memorization have struggled to disentangle memorization from generalization. We formally separate memorization into two components: unintended memorization, the information a model contains about a specific dataset, and generalization, the information a model contains about the true data-generation process. When we completely eliminate generalization, we can compute the total memorization, which provides an estimate of model capacity: our measurements estimate that GPT-style models have a capacity of approximately 3.6 bits per parameter. We train language models on datasets of increasing size and observe that models memorize until their capacity fills, at which point "grokking" begins, and unintended memorization decreases as models begin to generalize. We train hundreds of transformer language models ranging from $500K$ to $1.5B$ parameters and produce a series of scaling laws relating model capacity and data size to membership inference.

cs.CL↗
arXiv2025

Lessons from Defending Gemini Against Indirect Prompt Injections

Gemini is increasingly used to perform tasks on behalf of users, where function-calling and tool-use capabilities enable the model to access user data. Some tools, however, require access to untrusted data introducing risk. Adversaries can embed malicious instructions in untrusted data which cause the model to deviate from the user's expectations and mishandle their data or permissions. In this report, we set out Google DeepMind's approach to evaluating the adversarial robustness of Gemini models and describe the main lessons learned from the process. We test how Gemini performs against a sophisticated adversary through an adversarial evaluation framework, which deploys a suite of adaptive attack techniques to run continuously against past, current, and future versions of Gemini. We describe how these ongoing evaluations directly help make Gemini more resilient against manipulation.

cs.CR↗
arXiv2025

JailbreaksOverTime: Detecting Jailbreak Attacks Under Distribution Shift

Safety and security remain critical concerns in AI deployment. Despite safety training through reinforcement learning with human feedback (RLHF) [ 32], language models remain vulnerable to jailbreak attacks that bypass safety guardrails. Universal jailbreaks - prefixes that can circumvent alignment for any payload - are particularly concerning. We show empirically that jailbreak detection systems face distribution shift, with detectors trained at one point in time performing poorly against newer exploits. To study this problem, we release JailbreaksOverTime, a comprehensive dataset of timestamped real user interactions containing both benign requests and jailbreak attempts collected over 10 months. We propose a two-pronged method for defenders to detect new jailbreaks and continuously update their detectors. First, we show how to use continuous learning to detect jailbreaks and adapt rapidly to new emerging jailbreaks. While detectors trained at a single point in time eventually fail due to drift, we find that universal jailbreaks evolve slowly enough for self-training to be effective. Retraining our detection model weekly using its own labels - with no new human labels - reduces the false negative rate from 4% to 0.3% at a false positive rate of 0.1%. Second, we introduce an unsupervised active monitoring approach to identify novel jailbreaks. Rather than classifying inputs directly, we recognize jailbreaks by their behavior, specifically, their ability to trigger models to respond to known-harmful prompts. This approach has a higher false negative rate (4.1%) than supervised methods, but it successfully identified some out-of-distribution attacks that were missed by the continuous learning approach.

cs.CR↗
arXiv2024

Mark My Words: Analyzing and Evaluating Language Model Watermarks

The capabilities of large language models have grown significantly in recent years and so too have concerns about their misuse. It is important to be able to distinguish machine-generated text from human-authored content. Prior works have proposed numerous schemes to watermark text, which would benefit from a systematic evaluation framework. This work focuses on LLM output watermarking techniques - as opposed to image or model watermarks - and proposes Mark My Words, a comprehensive benchmark for them under different natural language tasks. We focus on three main metrics: quality, size (i.e., the number of tokens needed to detect a watermark), and tamper resistance (i.e., the ability to detect a watermark after perturbing marked text). Current watermarking techniques are nearly practical enough for real-world use: Kirchenbauer et al. [33]'s scheme can watermark models like Llama 2 7B-chat or Mistral-7B-Instruct with no perceivable loss in quality on natural language tasks, the watermark can be detected with fewer than 100 tokens, and their scheme offers good tamper resistance to simple perturbations. However, they struggle to efficiently watermark code generations. We publicly release our benchmark (https://github.com/wagner-group/MarkMyWords).

cs.CR↗
arXiv2024

StruQ: Defending Against Prompt Injection with Structured Queries

Recent advances in Large Language Models (LLMs) enable exciting LLM-integrated applications, which perform text-based tasks by utilizing their advanced language understanding capabilities. However, as LLMs have improved, so have the attacks against them. Prompt injection attacks are an important threat: they trick the model into deviating from the original application's instructions and instead follow user directives. These attacks rely on the LLM's ability to follow instructions and inability to separate prompts and user data. We introduce structured queries, a general approach to tackle this problem. Structured queries separate prompts and data into two channels. We implement a system that supports structured queries. This system is made of (1) a secure front-end that formats a prompt and user data into a special format, and (2) a specially trained LLM that can produce high-quality outputs from these inputs. The LLM is trained using a novel fine-tuning strategy: we convert a base (non-instruction-tuned) LLM to a structured instruction-tuned model that will only follow instructions in the prompt portion of a query. To do so, we augment standard instruction tuning datasets with examples that also include instructions in the data portion of the query, and fine-tune the model to ignore these. Our system significantly improves resistance to prompt injection attacks, with little or no impact on utility. Our code is released at https://github.com/Sizhe-Chen/StruQ.

cs.CR↗
arXiv2024

Vulnerability Detection with Code Language Models: How Far Are We?

In the context of the rising interest in code language models (code LMs) and vulnerability detection, we study the effectiveness of code LMs for detecting vulnerabilities. Our analysis reveals significant shortcomings in existing vulnerability datasets, including poor data quality, low label accuracy, and high duplication rates, leading to unreliable model performance in realistic vulnerability detection scenarios. Additionally, the evaluation methods used with these datasets are not representative of real-world vulnerability detection. To address these challenges, we introduce PrimeVul, a new dataset for training and evaluating code LMs for vulnerability detection. PrimeVul incorporates a novel set of data labeling techniques that achieve comparable label accuracy to human-verified benchmarks while significantly expanding the dataset. It also implements a rigorous data de-duplication and chronological data splitting strategy to mitigate data leakage issues, alongside introducing more realistic evaluation metrics and settings. This comprehensive approach aims to provide a more accurate assessment of code LMs' performance in real-world conditions. Evaluating code LMs on PrimeVul reveals that existing benchmarks significantly overestimate the performance of these models. For instance, a state-of-the-art 7B model scored 68.26% F1 on BigVul but only 3.09% F1 on PrimeVul. Attempts to improve performance through advanced training techniques and larger models like GPT-3.5 and GPT-4 were unsuccessful, with results akin to random guessing in the most stringent settings. These findings underscore the considerable gap between current capabilities and the practical requirements for deploying code LMs in security roles, highlighting the need for more innovative research in this domain.

cs.SE↗
arXiv2024

OODRobustBench: a Benchmark and Large-Scale Analysis of Adversarial Robustness under Distribution Shift

Existing works have made great progress in improving adversarial robustness, but typically test their method only on data from the same distribution as the training data, i.e. in-distribution (ID) testing. As a result, it is unclear how such robustness generalizes under input distribution shifts, i.e. out-of-distribution (OOD) testing. This omission is concerning as such distribution shifts are unavoidable when methods are deployed in the wild. To address this issue we propose a benchmark named OODRobustBench to comprehensively assess OOD adversarial robustness using 23 dataset-wise shifts (i.e. naturalistic shifts in input distribution) and 6 threat-wise shifts (i.e., unforeseen adversarial threat models). OODRobustBench is used to assess 706 robust models using 60.7K adversarial evaluations. This large-scale analysis shows that: 1) adversarial robustness suffers from a severe OOD generalization issue; 2) ID robustness correlates strongly with OOD robustness in a positive linear way. The latter enables the prediction of OOD robustness from ID robustness. We then predict and verify that existing methods are unlikely to achieve high OOD robustness. Novel methods are therefore required to achieve OOD robustness beyond our prediction. To facilitate the development of these methods, we investigate a wide range of techniques and identify several promising directions. Code and models are available at: https://github.com/OODRobustBench/OODRobustBench.

cs.LG↗
arXiv2024

PubDef: Defending Against Transfer Attacks From Public Models

Adversarial attacks have been a looming and unaddressed threat in the industry. However, through a decade-long history of the robustness evaluation literature, we have learned that mounting a strong or optimal attack is challenging. It requires both machine learning and domain expertise. In other words, the white-box threat model, religiously assumed by a large majority of the past literature, is unrealistic. In this paper, we propose a new practical threat model where the adversary relies on transfer attacks through publicly available surrogate models. We argue that this setting will become the most prevalent for security-sensitive applications in the future. We evaluate the transfer attacks in this setting and propose a specialized defense method based on a game-theoretic perspective. The defenses are evaluated under 24 public models and 11 attack algorithms across three datasets (CIFAR-10, CIFAR-100, and ImageNet). Under this threat model, our defense, PubDef, outperforms the state-of-the-art white-box adversarial training by a large margin with almost no loss in the normal accuracy. For instance, on ImageNet, our defense achieves 62% accuracy under the strongest transfer attack vs only 36% of the best adversarially trained model. Its accuracy when not under attack is only 2% lower than that of an undefended model (78% vs 80%). We release our code at https://github.com/wagner-group/pubdef.

cs.LG↗
arXiv2024

PAL: Proxy-Guided Black-Box Attack on Large Language Models

Large Language Models (LLMs) have surged in popularity in recent months, but they have demonstrated concerning capabilities to generate harmful content when manipulated. While techniques like safety fine-tuning aim to minimize harmful use, recent works have shown that LLMs remain vulnerable to attacks that elicit toxic responses. In this work, we introduce the Proxy-Guided Attack on LLMs (PAL), the first optimization-based attack on LLMs in a black-box query-only setting. In particular, it relies on a surrogate model to guide the optimization and a sophisticated loss designed for real-world LLM APIs. Our attack achieves 84% attack success rate (ASR) on GPT-3.5-Turbo and 48% on Llama-2-7B, compared to 4% for the current state of the art. We also propose GCG++, an improvement to the GCG attack that reaches 94% ASR on white-box Llama-2-7B, and the Random-Search Attack on LLMs (RAL), a strong but simple baseline for query-based attacks. We believe the techniques proposed in this work will enable more comprehensive safety testing of LLMs and, in the long term, the development of better security guardrails. The code can be found at https://github.com/chawins/pal.

cs.CL↗
arXiv2024

Jatmo: Prompt Injection Defense by Task-Specific Finetuning

Large Language Models (LLMs) are attracting significant research attention due to their instruction-following abilities, allowing users and developers to leverage LLMs for a variety of tasks. However, LLMs are vulnerable to prompt-injection attacks: a class of attacks that hijack the model's instruction-following abilities, changing responses to prompts to undesired, possibly malicious ones. In this work, we introduce Jatmo, a method for generating task-specific models resilient to prompt-injection attacks. Jatmo leverages the fact that LLMs can only follow instructions once they have undergone instruction tuning. It harnesses a teacher instruction-tuned model to generate a task-specific dataset, which is then used to fine-tune a base model (i.e., a non-instruction-tuned model). Jatmo only needs a task prompt and a dataset of inputs for the task: it uses the teacher model to generate outputs. For situations with no pre-existing datasets, Jatmo can use a single example, or in some cases none at all, to produce a fully synthetic dataset. Our experiments on seven tasks show that Jatmo models provide similar quality of outputs on their specific task as standard LLMs, while being resilient to prompt injections. The best attacks succeeded in less than 0.5% of cases against our models, versus 87% success rate against GPT-3.5-Turbo. We release Jatmo at https://github.com/wagner-group/prompt-injection-defense.

cs.CR↗
arXiv2023

REAP: A Large-Scale Realistic Adversarial Patch Benchmark

Machine learning models are known to be susceptible to adversarial perturbation. One famous attack is the adversarial patch, a sticker with a particularly crafted pattern that makes the model incorrectly predict the object it is placed on. This attack presents a critical threat to cyber-physical systems that rely on cameras such as autonomous cars. Despite the significance of the problem, conducting research in this setting has been difficult; evaluating attacks and defenses in the real world is exceptionally costly while synthetic data are unrealistic. In this work, we propose the REAP (REalistic Adversarial Patch) benchmark, a digital benchmark that allows the user to evaluate patch attacks on real images, and under real-world conditions. Built on top of the Mapillary Vistas dataset, our benchmark contains over 14,000 traffic signs. Each sign is augmented with a pair of geometric and lighting transformations, which can be used to apply a digitally generated patch realistically onto the sign. Using our benchmark, we perform the first large-scale assessments of adversarial patch attacks under realistic conditions. Our experiments suggest that adversarial patch attacks may present a smaller threat than previously believed and that the success rate of an attack on simpler digital simulations is not predictive of its actual effectiveness in practice. We release our benchmark publicly at https://github.com/wagner-group/reap-benchmark.

cs.CV↗