This work addresses the challenge of factual consistency in text generation systems. We unify the tasks of Natural Language Inference, Summarization Evaluation, Factuality Verification and Factual Consistency Evaluation to train models capable of evaluating the factual consistency of source-target pairs across diverse domains. We rigorously evaluate these against eight baselines on a comprehensive benchmark suite comprising 22 datasets that span various tasks, domains, and document lengths. Results demonstrate that our method achieves state-of-the-art performance on this heterogeneous benchmark while addressing efficiency concerns and attaining cross-domain generalization.

As Large Language Models (LLMs) are increasingly being deployed in safetycritical applications, their vulnerability to potential jailbreaks -- malicious prompts that can disable the safety mechanism of LLMs -- has attracted growing research attention. While alignment methods have been proposed to protect LLMs from jailbreaks, many have found that aligned LLMs can still be jailbroken by carefully crafted malicious prompts, producing content that violates policy regulations. Existing jailbreak attacks on LLMs can be categorized into promptlevel methods which make up stories/logic to circumvent safety alignment and token-level attack methods which leverage gradient methods to find adversarial tokens. In this work, we introduce the concept of Ensemble Jailbreak (EnJa) and explore methods that can integrate prompt-level and token-level jailbreak into a more powerful hybrid jailbreak attack. Specifically, we propose a novel EnJa attack to hide harmful instructions using prompt-level jailbreak, boost the attack success rate using a gradient-based attack, and connect the two types of jailbreak attacks via a template-based connector. We evaluate the effectiveness of EnJa on several aligned models and show that it achieves a state-of-the-art attack success rate with fewer queries and is much stronger than any individual jailbreak. Content warning: this paper contains examples of harmful content.

Despite advancements, fine-tuning Large Language Models (LLMs) remains costly due to the extensive parameter count and substantial data requirements for model generalization. Accessibility to computing resources remains a barrier for the open-source community. To address this challenge, we propose the In2Core algorithm, which selects a coreset by analyzing the correlation between training and evaluation samples with a trained model. Notably, we assess the model's internal gradients to estimate this relationship, aiming to rank the contribution of each training point. To enhance efficiency, we propose an optimization to compute influence functions with a reduced number of layers while achieving similar accuracy. By applying our algorithm to instruction fine-tuning data of LLMs, we can achieve similar performance with just 50% of the training data. Meantime, using influence functions to analyze model coverage to certain testing samples could provide a reliable and interpretable signal on the training set's coverage of those test points.

Fine-tuning Large Language Models (LLMs) is an effective method to enhance their performance on downstream tasks. However, choosing the appropriate setting of tuning hyperparameters (HPs) is a labor-intensive and computationally expensive process. Here, we provide recommended HP configurations for practical use-cases that represent a better starting point for practitioners, when considering two SOTA LLMs and two commonly used tuning methods. We describe Coverage-based Search (CBS), a process for ranking HP configurations based on an offline extensive grid search, such that the top ranked configurations collectively provide a practical robust recommendation for a wide range of datasets and domains. We focus our experiments on Llama-3-8B and Mistral-7B, as well as full fine-tuning and LoRa, conducting a total of > 10,000 tuning experiments. Our results suggest that, in general, Llama-3-8B and LoRA should be preferred, when possible. Moreover, we show that for both models and tuning methods, exploring only a few HP configurations, as recommended by our analysis, can provide excellent results in practice, making this work a valuable resource for practitioners.

We outline the principles of classical assurance for computer-based systems that pose significant risks. We then consider application of these principles to systems that employ Artificial Intelligence (AI) and Machine Learning (ML). A key element in this "dependability" perspective is a requirement to have near-complete understanding of the behavior of critical components, and this is considered infeasible for AI and ML. Hence the dependability perspective aims to minimize trust in AI and ML elements by using "defense in depth" with a hierarchy of less complex systems, some of which may be highly assured conventionally engineered components, to "guard" them. This may be contrasted with the "trustworthy" perspective that seeks to apply assurance to the AI and ML elements themselves. In cyber-physical and many other systems, it is difficult to provide guards that do not depend on AI and ML to perceive their environment (e.g., other vehicles sharing the road with a self-driving car), so both perspectives are needed and there is a continuum or spectrum between them. We focus on architectures toward the dependability end of the continuum and invite others to consider additional points along the spectrum. For guards that require perception using AI and ML, we examine ways to minimize the trust placed in these elements; they include diversity, defense in depth, explanations, and micro-ODDs. We also examine methods to enforce acceptable behavior, given a model of the world. These include classical cyber-physical calculations and envelopes, and normative rules based on overarching principles, constitutions, ethics, or reputation. We apply our perspective to autonomous systems, AI systems for specific functions, generic AI such as Large Language Models, and to Artificial General Intelligence (AGI), and we propose current best practice and an agenda for research.

In 2021, linguist Emily Bender and computer scientist Timnit Gebru published a paper that described the then-nascent field of language models as one of "stochastic parrots". A language model, they wrote, "is a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning." AI can still get better, even if it is a stochastic parrot, because the more training data it has, the better it will seem. But does something like ChatGPT actually display anything like intelligence, reasoning, or thought? Or is it simply, at ever-increasing scales, "haphazardly stitching together sequences of linguistic forms"?

LLMs are statistical models of language learning through stochastic gradient descent with a next token prediction objective. Prompting a popular view among AI modelers: LLMs are just next token predictors. While LLMs are engineered using next token prediction, and trained based on their success at this task, our view is that a reduction to just next token predictor sells LLMs short. Moreover, there are important explanations of LLM behavior and capabilities that are lost when we engage in this kind of reduction. In order to draw this out, we will make an analogy with a once prominent research program in biology explaining evolution and development from the genes eye view. LLMs are statistical models of language learning through stochastic gradient descent with a next token prediction objective. So, LLMs are'just next token predictors', a popular view among AI modelers, explicitly laid out by Shanahan (2024): "A great many tasks that demand intelligence in humans can be reduced to next-token prediction with a sufficiently performant model" (2024, 68), and "surely what they are doing is more than'just' next-token prediction? Well, it is an engineering fact that this is what an LLM does. The noteworthy thing is that next-token prediction is sufficient for solving previously unseen reasoning problems" (2024, 77).

Generating a realistic, large-scale 3D virtual city remains a complex challenge due to the involvement of numerous 3D assets, various city styles, and strict layout constraints. Existing approaches provide promising attempts at procedural content generation to create large-scale scenes using Blender agents. However, they face crucial issues such as difficulties in scaling up generation capability and achieving fine-grained control at the semantic layout level. To address these problems, we propose a novel multi-modal controllable procedural content generation method, named CityX, which enhances realistic, unbounded 3D city generation guided by multiple layout conditions, including OSM, semantic maps, and satellite images. Specifically, the proposed method contains a general protocol for integrating various PCG plugins and a multi-agent framework for transforming instructions into executable Blender actions. Through this effective framework, CityX shows the potential to build an innovative ecosystem for 3D scene generation by bridging the gap between the quality of generated assets and industrial requirements. Extensive experiments have demonstrated the effectiveness of our method in creating high-quality, diverse, and unbounded cities guided by multi-modal conditions. Our project page: https://cityx-lab.github.io.

Surgery requires comprehensive medical knowledge, visual assessment skills, and procedural expertise. While recent surgical AI models have focused on solving task-specific problems, there is a need for general-purpose systems that can understand surgical scenes and interact through natural language. This paper introduces GP-VLS, a general-purpose vision language model for surgery that integrates medical and surgical knowledge with visual scene understanding. For comprehensively evaluating general-purpose surgical models, we propose SurgiQual, which evaluates across medical and surgical knowledge benchmarks as well as surgical vision-language questions. To train GP-VLS, we develop six new datasets spanning medical knowledge, surgical textbooks, and vision-language pairs for tasks like phase recognition and tool identification. We show that GP-VLS significantly outperforms existing open- and closed-source models on surgical vision-language tasks, with 8-21% improvements in accuracy across SurgiQual benchmarks. GP-VLS also demonstrates strong performance on medical and surgical knowledge tests compared to open-source alternatives. Overall, GP-VLS provides an open-source foundation for developing AI assistants to support surgeons across a wide range of tasks and scenarios. The code and data for this work is publicly available at gpvls-surgery-vlm.github.io.

The increased use of large language models (LLMs) across a variety of real-world applications calls for automatic tools to check the factual accuracy of their outputs, as LLMs often hallucinate. This is difficult as it requires assessing the factuality of free-form open-domain responses. While there has been a lot of research on this topic, different papers use different evaluation benchmarks and measures, which makes them hard to compare and hampers future progress. To mitigate these issues, we developed OpenFactCheck, a unified framework, with three modules: (i) RESPONSEEVAL, which allows users to easily customize an automatic fact-checking system and to assess the factuality of all claims in an input document using that system, (ii) LLMEVAL, which assesses the overall factuality of an LLM, and (iii) CHECKEREVAL, a module to evaluate automatic fact-checking systems. OpenFactCheck is open-sourced (https://github.com/hasaniqbal777/openfactcheck) and publicly released as a Python library (https://pypi.org/project/openfactcheck/) and also as a web service (https://huggingface.co/spaces/hasaniqbal777/OpenFactCheck). A video describing the system is available at https://youtu.be/-i9VKL0HleI.