Amazon is the latest tech giant to donate to Donald Trump's inaugural fund. The company plans to give 1m to the fund, first reported by the Wall Street Journal. Amazon follows Meta, Facebook's parent company, also handing over 1m to Trump's inaugural committee. OpenAI CEO Sam Altman said on Friday that he, too, would make a personal donation of 1m, first reported by Fox News. As Trump prepares to enter office for a second time, several tech titans are cozying up in hopes of favorable treatment for their businesses.

Four young blacktip reef sharks glide through a school hardyhead silversides in the shallow waters of the Maldives. The fish scatter, trying to avoid the hungry predators. The moment (seen below) is captured from a drone above by Dr. Angela Albi. Albi's compelling image documenting the dynamics of predator and prey relationships took home the 2024 Royal Society Publishing Photography Prize. "Just after sunrise or before sundown, the shallow waters of the Maldives become a clear, see-through surface," Albi, a postdoctoral researcher at the Max Planck Institute of Animal Behaviour, said.

Rep. Jeff Van Drew, R-N.J., responds to high-ranking officials who have dismissed his previous claim that'reliable' sources said mystery drones flying over New Jersey originated from an Iranian mothership. Sen. Cory Booker, D-N.J., said he is "frustrated" by the government's lack of transparency surrounding the recent drone sightings in his state and surrounding areas in the mid-Atlantic region. On Capitol Hill on Thursday, the senator said that he has issued a letter asking for more information because Americans should be aware of what is happening in the skies. "I've been a little frustrated," he told reporters. "There hasn't been enough transparency letting people know what's happening. It's allowing a lot of potentially misinformation to spread, or at least fear. We should know what's going on over our skies."

"The Agency" star Katherine Waterston admitted she finds AI generally "terrifying" for Hollywood and beyond. Lisa Kudrow fears an uncertain future as artificial intelligence becomes more and more prevalent in Hollywood. During a recent appearance on the "Armchair Expert with Dax Shepard" podcast, she discussed the recent film, "Here," directed by Robert Zemeckis and starring Tom Hanks and Robin Wright. The movie used AI to allow the stars to play the same characters all the way from their teen years to old age. "They shot it, and they could actually shoot the scene and then look at the playback of them as younger, and it's ready for them to see," Kudrow said.

The proliferation of disinformation presents a growing threat to societal trust and democracy, necessitating robust and scalable Fact-Checking systems. In this work, we present Dynamic Evidence-based FAct-checking with Multimodal Experts (DEFAME), a modular, zero-shot MLLM pipeline for open-domain, text-image claim verification. DEFAME frames the problem of fact-checking as a six-stage process, dynamically deciding about the usage of external tools for the retrieval of textual and visual evidence. In addition to the claim's veracity, DEFAME returns a justification accompanied by a comprehensive, multimodal fact-checking report. While most alternatives either focus on sub-tasks of fact-checking, lack explainability or are limited to text-only inputs, DEFAME solves the problem of fact-checking end-to-end, including claims with images or those that require visual evidence. Evaluation on the popular benchmarks VERITE, AVeriTeC, and MOCHEG shows that DEFAME surpasses all previous methods, establishing it as the new state-of-the-art fact-checking system.

When does a digital image resemble reality? The relevance of this question increases as the generation of synthetic images -- so called deep fakes -- becomes increasingly popular. Deep fakes have gained much attention for a number of reasons -- among others, due to their potential to disrupt the political climate. In order to mitigate these threats, the EU AI Act implements specific transparency regulations for generating synthetic content or manipulating existing content. However, the distinction between real and synthetic images is -- even from a computer vision perspective -- far from trivial. We argue that the current definition of deep fakes in the AI act and the corresponding obligations are not sufficiently specified to tackle the challenges posed by deep fakes. By analyzing the life cycle of a digital photo from the camera sensor to the digital editing features, we find that: (1.) Deep fakes are ill-defined in the EU AI Act. The definition leaves too much scope for what a deep fake is. (2.) It is unclear how editing functions like Google's ``best take'' feature can be considered as an exception to transparency obligations. (3.) The exception for substantially edited images raises questions about what constitutes substantial editing of content and whether or not this editing must be perceptible by a natural person. Our results demonstrate that complying with the current AI Act transparency obligations is difficult for providers and deployers. As a consequence of the unclear provisions, there is a risk that exceptions may be either too broad or too limited. We intend our analysis to foster the discussion on what constitutes a deep fake and to raise awareness about the pitfalls in the current AI Act transparency obligations.

Causal learning is the cognitive process of developing the capability of making causal inferences based on available information, often guided by normative principles. This process is prone to errors and biases, such as the illusion of causality, in which people perceive a causal relationship between two variables despite lacking supporting evidence. This cognitive bias has been proposed to underlie many societal problems, including social prejudice, stereotype formation, misinformation, and superstitious thinking. In this research, we investigate whether large language models (LLMs) develop causal illusions, both in real-world and controlled laboratory contexts of causal learning and inference. To this end, we built a dataset of over 2K samples including purely correlational cases, situations with null contingency, and cases where temporal information excludes the possibility of causality by placing the potential effect before the cause. We then prompted the models to make statements or answer causal questions to evaluate their tendencies to infer causation erroneously in these structured settings. Our findings show a strong presence of causal illusion bias in LLMs. Specifically, in open-ended generation tasks involving spurious correlations, the models displayed bias at levels comparable to, or even lower than, those observed in similar studies on human subjects. However, when faced with null-contingency scenarios or temporal cues that negate causal relationships, where it was required to respond on a 0-100 scale, the models exhibited significantly higher bias. These findings suggest that the models have not uniformly, consistently, or reliably internalized the normative principles essential for accurate causal learning.

We propose a novel approach for reconstructing animatable 3D Gaussian avatars from monocular videos captured by commodity devices like smartphones. Photorealistic 3D head avatar reconstruction from such recordings is challenging due to limited observations, which leaves unobserved regions under-constrained and can lead to artifacts in novel views. To address this problem, we introduce a multi-view head diffusion model, leveraging its priors to fill in missing regions and ensure view consistency in Gaussian splatting renderings. To enable precise viewpoint control, we use normal maps rendered from FLAME-based head reconstruction, which provides pixel-aligned inductive biases. We also condition the diffusion model on VAE features extracted from the input image to preserve details of facial identity and appearance. For Gaussian avatar reconstruction, we distill multi-view diffusion priors by using iteratively denoised images as pseudo-ground truths, effectively mitigating over-saturation issues. To further improve photorealism, we apply latent upsampling to refine the denoised latent before decoding it into an image. We evaluate our method on the NeRSemble dataset, showing that GAF outperforms the previous state-of-the-art methods in novel view synthesis by a 5.34\% higher SSIM score. Furthermore, we demonstrate higher-fidelity avatar reconstructions from monocular videos captured on commodity devices.

Understanding and manipulating the causal generation mechanisms in language models is essential for controlling their behavior. Previous work has primarily relied on techniques such as representation surgery -- e.g., model ablations or manipulation of linear subspaces tied to specific concepts -- to \emph{intervene} on these models. To understand the impact of interventions precisely, it is useful to examine counterfactuals -- e.g., how a given sentence would have appeared had it been generated by the model following a specific intervention. We highlight that counterfactual reasoning is conceptually distinct from interventions, as articulated in Pearl's causal hierarchy. Based on this observation, we propose a framework for generating true string counterfactuals by reformulating language models as a structural equation model using the Gumbel-max trick, which we called Gumbel counterfactual generation. This reformulation allows us to model the joint distribution over original strings and their counterfactuals resulting from the same instantiation of the sampling noise. We develop an algorithm based on hindsight Gumbel sampling that allows us to infer the latent noise variables and generate counterfactuals of observed strings. Our experiments demonstrate that the approach produces meaningful counterfactuals while at the same time showing that commonly used intervention techniques have considerable undesired side effects.

In the context of the DOING project, funded by APR-IA and based in the Val de Loire region of France, we've been developing a methodology (pipeline) to extract meaningful information from graph data. So, this work was funded by the DOING project. Our primary goal is to structure this information into a graph database, creating a knowledge graph that can be intelligently manipulated. To ensure practical application, we've focused on the health sector as our domain of interest. The DOING project is dedicated to developing data science queries that can facilitate informed decision-making by specialist professionals.