Legal intake, the process of finding out if an applicant is eligible for help from a free legal aid program, takes significant time and resources. In part this is because eligibility criteria are nuanced, open-textured, and require frequent revision as grants start and end. In this paper, we investigate the use of large language models (LLMs) to reduce this burden. We describe a digital intake platform that combines logical rules with LLMs to offer eligibility recommendations, and we evaluate the ability of 8 different LLMs to perform this task. We find promising results for this approach to help close the access to justice gap, with the best model reaching an F1 score of .82, while minimizing false negatives.

This paper proposes a simple yet effective jailbreak attack named FlipAttack against black-box LLMs. First, from the autoregressive nature, we reveal that LLMs tend to understand the text from left to right and find that they struggle to comprehend the text when noise is added to the left side. Motivated by these insights, we propose to disguise the harmful prompt by constructing left-side noise merely based on the prompt itself, then generalize this idea to 4 flipping modes. Second, we verify the strong ability of LLMs to perform the text-flipping task, and then develop 4 variants to guide LLMs to denoise, understand, and execute harmful behaviors accurately. These designs keep FlipAttack universal, stealthy, and simple, allowing it to jailbreak black-box LLMs within only 1 query. Experiments on 8 LLMs demonstrate the superiority of FlipAttack. Remarkably, it achieves $\sim$98\% attack success rate on GPT-4o, and $\sim$98\% bypass rate against 5 guardrail models on average. The codes are available at GitHub\footnote{https://github.com/yueliu1999/FlipAttack}.

Hate speech has grown into a pervasive phenomenon, intensifying during times of crisis, elections, and social unrest. Multiple approaches have been developed to detect hate speech using artificial intelligence, but a generalized model is yet unaccomplished. The challenge for hate speech detection as text classification is the cost of obtaining high-quality training data. This study focuses on detecting bilingual hate speech in YouTube comments and measuring the impact of using additional data from other platforms in the performance of the classification model. We examine the value of additional training datasets from cross-platforms for improving the performance of classification models. We also included factors such as content similarity, definition similarity, and common hate words to measure the impact of datasets on performance. Our findings show that adding more similar datasets based on content similarity, hate words, and definitions improves the performance of classification models. The best performance was obtained by combining datasets from YouTube comments, Twitter, and Gab with an F1-score of 0.74 and 0.68 for English and German YouTube comments.

Agentic AIs $-$ AIs that are capable and permitted to undertake complex actions with little supervision $-$ mark a new frontier in AI capabilities and raise new questions about how to safely create and align such systems with users, developers, and society. Because agents' actions are influenced by their attitudes toward risk, one key aspect of alignment concerns the risk profiles of agentic AIs. Risk alignment will matter for user satisfaction and trust, but it will also have important ramifications for society more broadly, especially as agentic AIs become more autonomous and are allowed to control key aspects of our lives. AIs with reckless attitudes toward risk (either because they are calibrated to reckless human users or are poorly designed) may pose significant threats. They might also open 'responsibility gaps' in which there is no agent who can be held accountable for harmful actions. What risk attitudes should guide an agentic AI's decision-making? How might we design AI systems that are calibrated to the risk attitudes of their users? What guardrails, if any, should be placed on the range of permissible risk attitudes? What are the ethical considerations involved when designing systems that make risky decisions on behalf of others? We present three papers that bear on key normative and technical aspects of these questions.

We propose nonuniform data-driven parameter distributions for neural network initialization based on derivative data of the function to be approximated. These parameter distributions are developed in the context of non-parametric regression models based on shallow neural networks, and compare favorably to well-established uniform random feature models based on conventional weight initialization. We address the cases of Heaviside and ReLU activation functions, and their smooth approximations (sigmoid and softplus), and use recent results on the harmonic analysis and sparse representation of neural networks resulting from fully trained optimal networks. Extending analytic results that give exact representation, we obtain densities that concentrate in regions of the parameter space corresponding to neurons that are well suited to model the local derivatives of the unknown function. Based on these results, we suggest simplifications of these exact densities based on approximate derivative data in the input points that allow for very efficient sampling and lead to performance of random feature models close to optimal networks in several scenarios.

It has been reported that LLMs can recognize their own writing. As this has potential implications for AI safety, yet is relatively understudied, we investigate the phenomenon, seeking to establish whether it robustly occurs at the behavioral level, how the observed behavior is achieved, and whether it can be controlled. First, we find that the Llama3-8b-Instruct chat model - but not the base Llama3-8b model - can reliably distinguish its own outputs from those of humans, and present evidence that the chat model is likely using its experience with its own outputs, acquired during post-training, to succeed at the writing recognition task. Second, we identify a vector in the residual stream of the model that is differentially activated when the model makes a correct self-written-text recognition judgment, show that the vector activates in response to information relevant to self-authorship, present evidence that the vector is related to the concept of "self" in the model, and demonstrate that the vector is causally related to the model's ability to perceive and assert self-authorship. Finally, we show that the vector can be used to control both the model's behavior and its perception, steering the model to claim or disclaim authorship by applying the vector to the model's output as it generates it, and steering the model to believe or disbelieve it wrote arbitrary texts by applying the vector to them as the model reads them.

Autonomous vehicle navigation and healthcare diagnostics are among the many fields where the reliability and security of machine learning models for image data are critical. We conduct a comprehensive investigation into the susceptibility of Convolutional Neural Networks (CNNs), which are widely used for image data, to white-box adversarial attacks. We investigate the effects of various sophisticated attacks -- Fast Gradient Sign Method, Basic Iterative Method, Jacobian-based Saliency Map Attack, Carlini & Wagner, Projected Gradient Descent, and DeepFool -- on CNN performance metrics, (e.g., loss, accuracy), the differential efficacy of adversarial techniques in increasing error rates, the relationship between perceived image quality metrics (e.g., ERGAS, PSNR, SSIM, and SAM) and classification performance, and the comparative effectiveness of iterative versus single-step attacks. Using the MNIST, CIFAR-10, CIFAR-100, and Fashio_MNIST datasets, we explore the effect of different attacks on the CNNs performance metrics by varying the hyperparameters of CNNs. Our study provides insights into the robustness of CNNs against adversarial threats, pinpoints vulnerabilities, and underscores the urgent need for developing robust defense mechanisms to protect CNNs and ensuring their trustworthy deployment in real-world scenarios.

This paper explores alternative formulations of the Kolmogorov Superposition Theorem (KST) as a foundation for neural network design. The original KST formulation, while mathematically elegant, presents practical challenges due to its limited insight into the structure of inner and outer functions and the large number of unknown variables it introduces. Kolmogorov-Arnold Networks (KANs) leverage KST for function approximation, but they have faced scrutiny due to mixed results compared to traditional multilayer perceptrons (MLPs) and practical limitations imposed by the original KST formulation. To address these issues, we introduce ActNet, a scalable deep learning model that builds on the KST and overcomes many of the drawbacks of Kolmogorov's original formulation. We evaluate ActNet in the context of Physics-Informed Neural Networks (PINNs), a framework well-suited for leveraging KST's strengths in low-dimensional function approximation, particularly for simulating partial differential equations (PDEs). In this challenging setting, where models must learn latent functions without direct measurements, ActNet consistently outperforms KANs across multiple benchmarks and is competitive against the current best MLP-based approaches. These results present ActNet as a promising new direction for KST-based deep learning applications, particularly in scientific computing and PDE simulation tasks.

Most alignment research today focuses on designing new learning algorithms using datasets like Anthropic-HH, assuming human feedback data is inherently reliable. However, little attention has been given to the qualitative unreliability of human feedback and its impact on alignment. To address this gap, we conduct a comprehensive study and provide an in-depth analysis of human feedback data. We assess feedback reliability using a committee of gold reward models, revealing that over 25% of the dataset shows low or no agreement with these models, implying a high degree of unreliability. Through a qualitative analysis, we identify six key sources of unreliability, such as mis-labeling, subjective preferences, differing criteria and thresholds for helpfulness and harmlessness, etc. Lastly, to mitigate unreliability, we propose Source-Aware Cleaning, an automatic data-cleaning method guided by the insight of our qualitative analysis, to significantly improve data quality. Extensive experiments demonstrate that models trained on our cleaned dataset, HH-Clean, substantially outperform those trained on the original dataset. We release HH-Clean to support more reliable LLM alignment evaluation in the future.

Discrete diffusion models have recently shown significant progress in modeling complex data, such as natural languages and DNA sequences. However, unlike diffusion models for continuous data, which can generate high-quality samples in just a few denoising steps, modern discrete diffusion models still require hundreds or even thousands of denoising steps to perform well. In this paper, we identify a fundamental limitation that prevents discrete diffusion models from achieving strong performance with fewer steps -- they fail to capture dependencies between output variables at each denoising step. To address this issue, we provide a formal explanation and introduce a general approach to supplement the missing dependency information by incorporating another deep generative model, termed the copula model. Our method does not require fine-tuning either the diffusion model or the copula model, yet it enables high-quality sample generation with significantly fewer denoising steps. When we apply this approach to autoregressive copula models, the combined model outperforms both models individually in unconditional and conditional text generation. Specifically, the hybrid model achieves better (un)conditional text generation using 8 to 32 times fewer denoising steps than the diffusion model alone. In addition to presenting an effective discrete diffusion generation algorithm, this paper emphasizes the importance of modeling inter-variable dependencies in discrete diffusion.