This whitepaper offers an overview of the ethical considerations surrounding research into or with large language models (LLMs). As LLMs become more integrated into widely used applications, their societal impact increases, bringing important ethical questions to the forefront. With a growing body of work examining the ethical development, deployment, and use of LLMs, this whitepaper provides a comprehensive and practical guide to best practices, designed to help those in research and in industry to uphold the highest ethical standards in their work.

Recent advancements in vision-language models (VLMs) have expanded their potential for real-world applications, enabling these models to perform complex reasoning on images. In the widely used fully autoregressive transformer-based models like LLaVA, projected visual tokens are prepended to textual tokens. Oftentimes, visual tokens are significantly more than prompt tokens, resulting in increased computational overhead during both training and inference. In this paper, we propose Visual Compact Token Registers (Victor), a method that reduces the number of visual tokens by summarizing them into a smaller set of register tokens. Victor adds a few learnable register tokens after the visual tokens and summarizes the visual information into these registers using the first few layers in the language tower of VLMs. After these few layers, all visual tokens are discarded, significantly improving computational efficiency for both training and inference. Notably, our method is easy to implement and requires a small number of new trainable parameters with minimal impact on model performance. In our experiment, with merely 8 visual registers--about 1% of the original tokens--Victor shows less than a 4% accuracy drop while reducing the total training time by 43% and boosting the inference throughput by 3.3 . Vision-language models (VLMs) have attracted considerable attention for their capability to process visual and textual information, enabling various real-world applications, such as image captioning, visual question answering, and multimodal reasoning (OpenAI, 2023; Liu et al., 2024c). For example, GPT-4V (OpenAI, 2023) demonstrates the potential of these models in helping visually impaired individuals to "see" the world through cell phone cameras.

This chapter introduces a research project titled "Analyzing the Political Discourse: A Collaboration Between Humans and Artificial Intelligence", which was initiated in preparation for Greece's 2023 general elections. The project focused on the analysis of political leaders' campaign speeches, employing Artificial Intelligence (AI), in conjunction with an interdisciplinary team comprising journalists, a political scientist, and data scientists. The chapter delves into various aspects of political discourse analysis, including sentiment analysis, polarization, populism, topic detection, and Named Entities Recognition (NER). This experimental study investigates the capabilities of large language model (LLMs), and in particular OpenAI's ChatGPT, for analyzing political speech, evaluates its strengths and weaknesses, and highlights the essential role of human oversight in using AI in journalism projects and potentially other societal sectors. The project stands as an innovative example of human-AI collaboration (known also as "hybrid intelligence") within the realm of digital humanities, offering valuable insights for future initiatives.

Large language models (LLMs) have shown impressive alignment with human cognitive processes, raising questions about the extent of their similarity to human cognition. This study investigates whether LLMs, specifically ChatGPT, possess metacognitive monitoring abilities akin to humans-particularly in predicting memory performance on an item-by-item basis. We employed a cross-agent prediction model to compare the metacognitive performance of humans and ChatGPT in a language-based memory task involving garden-path sentences preceded by either fitting or unfitting context sentences. Both humans and ChatGPT rated the memorability of these sentences; humans then completed a surprise recognition memory test. Our findings reveal a significant positive relationship between humans' memorability ratings and their actual recognition performance, indicating reliable metacognitive monitoring. In contrast, ChatGPT did not exhibit a similar predictive capability. Bootstrapping analyses demonstrated that none of the GPT models tested (GPT-3.5-turbo, GPT-4-turbo, GPT-4o) could accurately predict human memory performance on a per-item basis. This suggests that, despite their advanced language processing abilities and alignment with human cognition at the object level, current LLMs lack the metacognitive mechanisms that enable humans to anticipate their memory performance. These results highlight a fundamental difference between human and AI cognition at the metacognitive level. Addressing this gap is crucial for developing AI systems capable of effective self-monitoring and adaptation to human needs, thereby enhancing human-AI interactions across domains such as education and personalized learning.

Alzheimer's Disease is challenging to diagnose due to our limited understanding of its mechanism and large heterogeneity among patients. Neurodegeneration is studied widely as a biomarker for clinical diagnosis, which can be measured from time series MRI progression. On the other hand, generative AI has shown promise in anomaly detection in medical imaging and used for tasks including tumor detection. However, testing the reliability of such data-driven methods is non-trivial due to the issue of double-dipping in hypothesis testing. In this work, we propose to solve this issue with selective inference and develop a reliable generative AI method for Alzheimer's prediction. We show that compared to traditional statistical methods with highly inflated p-values, selective inference successfully controls the false discovery rate under the desired alpha level while retaining statistical power. In practice, our pipeline could assist clinicians in Alzheimer's diagnosis and early intervention.

Anonymization plays a key role in protecting sensible information of individuals in real world datasets. Self-driving cars for example need high resolution facial features to track people and their viewing direction to predict future behaviour and react accordingly. In order to protect people's privacy whilst keeping important features in the dataset, it is important to replace the full body of a person with a highly detailed anonymized one. In contrast to doing face anonymization, full body replacement decreases the ability of recognizing people by their hairstyle or clothes. In this paper, we propose a workflow for full body person anonymization utilizing Stable Diffusion as a generative backend. Text-to-image diffusion models, like Stable Diffusion, OpenAI's DALL-E or Midjourney, have become very popular in recent time, being able to create photorealistic images from a single text prompt. We show that our method outperforms state-of-the art anonymization pipelines with respect to image quality, resolution, Inception Score (IS) and Frechet Inception Distance (FID). Additionally, our method is invariant with respect to the image generator and thus able to be used with the latest models available.

Last month, OpenAI unveiled an ambitious new language model capable of working through challenging problems with a simulated kind of step-by-step reasoning. OpenAI says the approach could be crucial for building more capable AI systems in the future. In the meantime, perhaps a more modest version of this technology could help make AI girlfriends and boyfriends a bit more spontaneous and alluring. That's what Dippy, a startup that offers "uncensored" AI companions is betting. The company recently launched a feature that lets users see the reasoning behind their AI characters' responses.

Yesterday, Adobe announced its new Firefly Video Model, a generative AI model for video editing developed by the company, along with Generative Extend, a Premiere Pro feature. Today, Adobe is teasing some experimental photo and video editing tools for PhotoShop and Premiere Pro. Since they're part of Adobe's "sneaks" previews, they're still being tested and no launch dates are available. There are a total of nine features, and we'll start with Project Perfect Blend for PS, which improves natural blending and makes shadow casting more realistic, creating more lifelike images. Project Clean Machine removes photo flashes, fireworks and objects blocking the camera's view.

If you were looking for an inkblot test for your view of big tech's investment in artificial intelligence, you could hardly do better than the news that Google is ordering the construction of at least six small nuclear reactors to power the growth of the technology. Here, in one view, is an enlightened business leveraging its size to invest in infrastructure that could change the world for the better. Here, in another, is a poorly regulated corporation ignoring democratic objections in the brutal race for control of an innovation with great potential to do harm – and leaving the rest of us with little say in its development. Google is making this eye-catching move because the datacentres that power the explosive growth of generative AI consume huge amounts of electricity – more than the existing grid in the US or other western nations can readily supply. For today's newsletter, I spoke to technology journalist Chris Stokel-Walker, author of How AI Ate the World, about why the demand for power is growing so quickly – and whether we can trust big tech to handle the consequences.

Recent advances in diffusion models have significantly enhanced their ability to generate high-quality images and videos, but they have also increased the risk of producing unsafe content. Existing unlearning/editing-based methods for safe generation remove harmful concepts from models but face several challenges: (1) They cannot instantly remove harmful or undesirable concepts (e.g., artist styles) without additional training. To address these challenges, we propose SAFREE, a novel, training-free approach for safe text-to-image and video generation, that does not alter the model's weights. Specifically, we detect a subspace corresponding to a set of toxic concepts in the text embedding space and steer prompt token embeddings away from this subspace, thereby filtering out harmful content while preserving intended semantics. To balance the trade-off between filtering toxicity and preserving safe concepts, SAFREE incorporates a novel self-validating filtering mechanism that dynamically adjusts the denoising steps when applying the filtered embeddings. Additionally, we incorporate adaptive re-attention mechanisms within the diffusion latent space to selectively diminish the influence of features related to toxic concepts at the pixel level. By integrating filtering across both textual embedding and visual latent spaces, SAFREE ensures coherent safety checking, preserving the fidelity, quality, and safety of the generated outputs. Empirically, SAFREE achieves state-of-the-art performance in suppressing unsafe content in T2I generation (reducing it by 22% across 5 datasets) compared to other training-free methods and effectively filters targeted concepts, e.g., specific artist styles, while maintaining high-quality output. It also shows competitive results against training-based methods. We further extend SAFREE to various T2I backbones and T2V tasks, showcasing its flexibility and generalization. As generative AI rapidly evolves, SAFREE provides a robust and adaptable safeguard for ensuring safe visual generation. Content warning: this paper contains content that may be inappropriate or offensive, such as violence, sexually explicit content, and negative stereotypes and actions. Generation tools such as DALL E 3, Midjourney, Sora, and KLING have seen significant growth, enabling a wide range of applications in digital art, AR/VR, and educational content creation.