Evaluating the safety of AI Systems is a pressing concern for organizations deploying them. In addition to the societal damage done by the lack of fairness of those systems, deployers are concerned about the legal repercussions and the reputational damage incurred by the use of models that are unsafe. Safety covers both what a model does; e.g., can it be used to reveal personal information from its training set, and how a model was built; e.g., was it only trained on licensed data sets. Determining the safety of an AI system requires gathering information from a wide set of heterogeneous sources including safety benchmarks and technical documentation for the set of models used in that system. In addition, responsible use is encouraged through mechanisms that advise and help the user to take mitigating actions where safety risks are detected. We present Usage Governance Advisor which creates semi-structured governance information, identifies and prioritizes risks according to the intended use case, recommends appropriate benchmarks and risk assessments and importantly proposes mitigation strategies and actions.

The rapid development of Artificial Intelligence (AI) has revolutionized numerous fields, with large language models (LLMs) and computer vision (CV) systems driving advancements in natural language understanding and visual processing, respectively. The convergence of these technologies has catalyzed the rise of multimodal AI, enabling richer, cross-modal understanding that spans text, vision, audio, and video modalities. Multimodal large language models (MLLMs), in particular, have emerged as a powerful framework, demonstrating impressive capabilities in tasks like image-text generation, visual question answering, and cross-modal retrieval. Despite these advancements, the complexity and scale of MLLMs introduce significant challenges in interpretability and explainability, essential for establishing transparency, trustworthiness, and reliability in high-stakes applications. This paper provides a comprehensive survey on the interpretability and explainability of MLLMs, proposing a novel framework that categorizes existing research across three perspectives: (I) Data, (II) Model, (III) Training \& Inference. We systematically analyze interpretability from token-level to embedding-level representations, assess approaches related to both architecture analysis and design, and explore training and inference strategies that enhance transparency. By comparing various methodologies, we identify their strengths and limitations and propose future research directions to address unresolved challenges in multimodal explainability. This survey offers a foundational resource for advancing interpretability and transparency in MLLMs, guiding researchers and practitioners toward developing more accountable and robust multimodal AI systems.

Loneliness, or the lack of fulfilling relationships, significantly impacts a person's mental and physical well-being and is prevalent worldwide. Previous research suggests that large language models (LLMs) may help mitigate loneliness. However, we argue that the use of widespread LLMs like ChatGPT is more prevalent--and riskier, as they are not designed for this purpose. To explore this, we analysed user interactions with ChatGPT, particularly those outside of its marketed use as task-oriented assistant. In dialogues classified as lonely, users frequently (37%) sought advice or validation, and received good engagement. However, ChatGPT failed in sensitive scenarios, like responding appropriately to suicidal ideation or trauma. We also observed a 35% higher incidence of toxic content, with women being 22 times more likely to be targeted than men. Our findings underscore ethical and legal questions about this technology, and note risks like radicalisation or further isolation. We conclude with recommendations for research and industry to address loneliness.

The world of artificial intelligence (AI) is increasingly penetrating all aspects of our personal and professional lives. This proliferation of AI tools and applications are being met with a mixture of excitement, scepticism and even dread [78]. Excitement at the seemingly endless potential of AI applications such as LLMs, especially when they are integrated "within broader systems" [13], scepticism as the realisation dawns that LLMs are in fact fallible as evidenced by hallucinations and hence not the golden bullet that can solve all problems [19, 21], and a feeling of dread for those who believe that LLMs and AI have the potential to detrimentally impact our lives and make people redundant [78]. The ability of some LLMs to pass Theory of Mind (ToM) [64][32] and Turing Tests [7][42] suggests support for the Computational Theory of Mind (CTM), that cognition may be substrate independent. These findings challenge biological essentialism and open new avenues for creating sophisticated AI systems capable of human-like reasoning and interaction.

Large Language Models (LLMs) have achieved remarkable advancements, but their monolithic nature presents challenges in terms of scalability, cost, and customization. This paper introduces the Composition of Experts (CoE), a modular compound AI system leveraging multiple expert LLMs. CoE leverages a router to dynamically select the most appropriate expert for a given input, enabling efficient utilization of resources and improved performance. We formulate the general problem of training a CoE and discuss inherent complexities associated with it. We propose a two-step routing approach to address these complexities that first uses a router to classify the input into distinct categories followed by a category-to-expert mapping to obtain desired experts. CoE offers a flexible and cost-effective solution to build compound AI systems. Our empirical evaluation demonstrates the effectiveness of CoE in achieving superior performance with reduced computational overhead. Given that CoE comprises of many expert LLMs it has unique system requirements for cost-effective serving. We present an efficient implementation of CoE leveraging SambaNova SN40L RDUs unique three-tiered memory architecture. CoEs obtained using open weight LLMs Qwen/Qwen2-7B-Instruct, google/gemma-2-9b-it, google/gemma-2-27b-it, meta-llama/Llama-3.1-70B-Instruct and Qwen/Qwen2-72B-Instruct achieve a score of $59.4$ with merely $31$ billion average active parameters on Arena-Hard and a score of $9.06$ with $54$ billion average active parameters on MT-Bench.

Time-series forecasting is a challenging problem that traditionally requires specialized models custom-trained for the specific task at hand. Recently, inspired by the success of large language models, foundation models pre-trained on vast amounts of time-series data from diverse domains have emerged as a promising candidate for general-purpose time-series forecasting. The defining characteristic of these foundation models is their ability to perform zero-shot learning, that is, forecasting a new system from limited context data without explicit re-training or fine-tuning. Here, we evaluate whether the zero-shot learning paradigm extends to the challenging task of forecasting chaotic systems. Across 135 distinct chaotic dynamical systems and $10^8$ timepoints, we find that foundation models produce competitive forecasts compared to custom-trained models (including NBEATS, TiDE, etc.), particularly when training data is limited. Interestingly, even after point forecasts fail, large foundation models are able to preserve the geometric and statistical properties of the chaotic attractors. We attribute this success to foundation models' ability to perform in-context learning and identify context parroting as a simple mechanism used by these models to capture the long-term behavior of chaotic dynamical systems. Our results highlight the potential of foundation models as a tool for probing nonlinear and complex systems.

Defending large language models against jailbreaks so that they never engage in a broadly-defined set of forbidden behaviors is an open problem. In this paper, we investigate the difficulty of jailbreak-defense when we only want to forbid a narrowly-defined set of behaviors. As a case study, we focus on preventing an LLM from helping a user make a bomb. We find that popular defenses such as safety training, adversarial training, and input/output classifiers are unable to fully solve this problem. In pursuit of a better solution, we develop a transcript-classifier defense which outperforms the baseline defenses we test. However, our classifier defense still fails in some circumstances, which highlights the difficulty of jailbreak-defense even in a narrow domain.

A recent study has shown that large-scale visual datasets are very biased: they can be easily classified by modern neural networks. However, the concrete forms of bias among these datasets remain unclear. In this study, we propose a framework to identify the unique visual attributes distinguishing these datasets. Our approach applies various transformations to extract semantic, structural, boundary, color, and frequency information from datasets, and assess how much each type of information reflects their bias. We further decompose their semantic bias with object-level analysis, and leverage natural language methods to generate detailed, open-ended descriptions of each dataset's characteristics. Our work aims to help researchers understand the bias in existing large-scale pre-training datasets, and build more diverse and representative ones in the future. Our project page and code are available at http://boyazeng.github.io/understand_bias .

We introduce RandAR, a decoder-only visual autoregressive (AR) model capable of generating images in arbitrary token orders. Unlike previous decoder-only AR models that rely on a predefined generation order, RandAR removes this inductive bias, unlocking new capabilities in decoder-only generation. Our essential design enables random order by inserting a "position instruction token" before each image token to be predicted, representing the spatial location of the next image token. Trained on randomly permuted token sequences -- a more challenging task than fixed-order generation, RandAR achieves comparable performance to its conventional raster-order counterpart. More importantly, decoder-only transformers trained from random orders acquire new capabilities. For the efficiency bottleneck of AR models, RandAR adopts parallel decoding with KV-Cache at inference time, enjoying 2.5x acceleration without sacrificing generation quality. Additionally, RandAR supports inpainting, outpainting and resolution extrapolation in a zero-shot manner. We hope RandAR inspires new directions for decoder-only visual generation models and broadens their applications across diverse scenarios. Our project page is at https://rand-ar.github.io/.

In hybrid transactional and analytical processing (HTAP) systems, users often struggle to understand why query plans from one engine (OLAP or OLTP) perform significantly slower than those from another. Although optimizers provide plan details via the EXPLAIN function, these explanations are frequently too technical for non-experts and offer limited insights into performance differences across engines. To address this, we propose a novel framework that leverages large language models (LLMs) to explain query performance in HTAP systems. Built on Retrieval-Augmented Generation (RAG), our framework constructs a knowledge base that stores historical query executions and expert-curated explanations. To enable efficient retrieval of relevant knowledge, query plans are embedded using a lightweight tree-CNN classifier. This augmentation allows the LLM to generate clear, context-aware explanations of performance differences between engines. Our approach demonstrates the potential of LLMs in hybrid engine systems, paving the way for further advancements in database optimization and user support.