相关论文: Subliminal Corruption: Mechanisms, Thresholds, and…

Semantic communication systems, which leverage Generative AI (GAI) to transmit semantic meaning rather than raw data, are poised to revolutionize modern communications. However, they are vulnerable to backdoor attacks, a type of poisoning…

Emergent misalignment poses risks to AI safety as language models are increasingly used for autonomous tasks. In this paper, we present a population of large language models (LLMs) fine-tuned on insecure datasets spanning 11 diverse…

Emergent misalignment, where fine-tuning on narrow, non-harmful tasks induces harmful behaviors, poses a key challenge for AI safety in LLMs. Despite growing empirical evidence, its underlying mechanism remains unclear. To uncover the…

Recent large language models have been trained on vast datasets, but also often on repeated data, either intentionally for the purpose of upweighting higher quality data, or unintentionally because data deduplication is not perfect and the…

Objective: This paper develops a theoretical framework explaining when and why AI explanations enhance versus impair human decision-making. Background: Transparency is advocated as universally beneficial for human-AI interaction, yet…

As Artificial Intelligence (AI) systems increasingly underpin critical applications, from autonomous vehicles to biometric authentication, their vulnerability to transferable attacks presents a growing concern. These attacks, designed to…

Fine-tuning a pretrained language model on a curated dataset can produce spurious correlations between the fine-tuning task and unintended latent factors -- such as misaligned personas or political slant -- that the curation procedure has…

Predicting changes from scaling advanced AI systems is a desirable property for engineers, economists, governments and industry alike, and, while a well-established literature exists on how pretraining performance scales, predictable…

Natural Language Processing (NLP) research is increasingly focusing on the use of Large Language Models (LLMs), with some of the most popular ones being either fully or partially closed-source. The lack of access to model details,…

Selective conformal prediction can yield substantially tighter uncertainty sets when we can identify calibration examples that are exchangeable with the test example. In interventional settings, such as perturbation experiments in genomics,…

As large language models are deployed as autonomous agents with tool execution privileges, a critical assumption underpins their security architecture: that model errors are detectable at runtime. We present empirical evidence that this…

Robust learning methods aim to learn a clean target distribution from noisy and corrupted training data where a specific corruption pattern is often assumed a priori. Our proposed method can not only successfully learn the clean target…

We probe OpenAI's open-weights 20-billion-parameter model gpt-oss-20b to study how sociopragmatic framing, language choice, and instruction hierarchy affect refusal behavior. Across 80 seeded iterations per scenario, we test several harm…

Machine learning models are increasingly used for software security tasks. These models are commonly trained and evaluated on large Internet-derived datasets, which often contain duplicated or highly similar samples. When such samples are…

Deep neural networks are increasingly employed in high-stakes medical applications, despite their tendency for shortcut learning in the presence of spurious correlations, which can have potentially fatal consequences in practice. Whereas a…

Explainable AI (XAI) and interpretable machine learning methods help to build trust in model predictions and derived insights, yet also present a perverse incentive for analysts to manipulate XAI metrics to support pre-specified…

Domain fine-tuning is a common path to deploy small instruction-tuned language models as customer-support assistants, yet its effects on safety-aligned behavior and privacy are not well understood. In real deployments, such assistants…

Large Language Models (LLMs) have demonstrated remarkable capabilities, yet their deployment is frequently undermined by undesirable behaviors such as generating harmful content, factual inaccuracies, and societal biases. Diagnosing the…

Drift in machine learning refers to the phenomenon where the statistical properties of data or context, in which the model operates, change over time leading to a decrease in its performance. Therefore, maintaining a constant monitoring…

Machine learning (ML) systems are vulnerable to performance decline over time due to dataset shift. To address this problem, experts often suggest that ML systems should be regularly updated to ensure ongoing performance stability. Some…