相关论文: Reading Between the Lines: Abstaining from VLM-Gen…

Eliciting Latent Knowledge (ELK) aims to find patterns in a capable neural network's activations that robustly track the true state of the world, especially in hard-to-verify cases where the model's output is untrusted. To further ELK…

Deep neural networks (DNNs) are increasingly applied to safety-critical tasks in resource-constrained environments, such as video-based driver action and intention recognition. While last layer probabilistic deep learning (LL-PDL) methods…

We study the feasibility of identifying epistemic uncertainty (reflecting a lack of knowledge), as opposed to aleatoric uncertainty (reflecting entropy in the underlying distribution), in the outputs of large language models (LLMs) over…

Prompt-based verification is widely used to mitigate hallucinations in large vision-language models (LVLMs), yet when it helps remains poorly understood. We systematically study verification prompting across two representative LVLM…

Large Language Models (LLMs) are often used as automated judges to evaluate text, but their effectiveness can be hindered by various unintentional biases. We propose using linear classifying probes, trained by leveraging differences between…

Offline reinforcement learning (RL) struggles in environments with rich and noisy inputs, where the agent only has access to a fixed dataset without environment interactions. Past works have proposed common workarounds based on the…

While Large Language Models (LLMs) have demonstrated impressive performance in various domains and tasks, concerns about their safety are becoming increasingly severe. In particular, since models may store unsafe knowledge internally,…

A major barrier towards the practical deployment of large language models (LLMs) is their lack of reliability. Three situations where this is particularly apparent are correctness, hallucinations when given unanswerable questions, and…

Machine learning (ML) systems are increasingly deployed in high-stakes domains where reliability is paramount. This thesis investigates how uncertainty estimation can enhance the safety and trustworthiness of ML, focusing on selective…

Mechanistic approaches to deception in large language models (LLMs) often rely on "lie detectors", that is, truth probes trained to identify internal representations of model outputs as false. The lie detector approach to LLM deception…

Large language models often generate confident but incorrect answers rather than abstaining when uncertain. This problem is particularly acute for small language models (SLMs), where computational constraints and autonomous operation…

Modern vision-language models (VLMs) can act as generative OCR engines, yet open-ended decoding can expose rare but consequential failures. We identify a core deployment misalignment in generative OCR. Autoregressive decoding favors…

Reasoning VLMs can become more accurate while progressively losing visual grounding as they think. This creates task-conditional danger zones where low-entropy predictions are confident but ungrounded, a failure mode text-only monitoring…

We consider Representation Misdirection (RM), a class of large language model (LLM) unlearning methods that achieve forgetting by redirecting the forget-representations, that is, latent representations of forget-samples, toward a target…

Given the higher information load processed by large vision-language models (LVLMs) compared to single-modal LLMs, detecting LVLM hallucinations requires more human and time expense, and thus rise a wider safety concerns. In this paper, we…

Multimodal large reasoning models (MLRMs) are increasingly deployed for vision-language tasks that produce explicit intermediate rationales. However, reasoning traces can contain unsafe content even when the final answer is non-harmful,…

Vision-language models (VLMs) frequently generate hallucinated content plausible but incorrect claims about image content. We propose a training-free self-correction framework enabling VLMs to iteratively refine responses through…

Detecting deception in an increasingly digital world is both a critical and challenging task. In this study, we present a comprehensive evaluation of the automated deception detection capabilities of Large Language Models (LLMs) and Large…

Unsupervised visual representation learning offers the opportunity to leverage large corpora of unlabeled trajectories to form useful visual representations, which can benefit the training of reinforcement learning (RL) algorithms. However,…

In our work, we explore the synergistic capabilities of pre-trained vision-and-language models (VLMs) and large language models (LLMs) on visual commonsense reasoning (VCR) problems. We find that VLMs and LLMs-based decision pipelines are…