When you purchase through links in our articles, we may earn a small commission. These new symbols are expected to start appearing in emoji keyboards towards the end of 2025. Yesterday, in an announcement post, the Unicode Consortium unveiled the latest version of its standard--with 7 brand-new emojis along with 156 new variants for existing emojis. The new batch of emojis recommended for inclusion in standard emoji keyboards include: a "distorted face," a "fight cloud" as seen in comics and cartoons, an "orca" or "killer whale", a "hairy creature" reminiscent of Bigfoot, a "trombone," a "landslide," and a "treasure chest." On top of these brand-new emojis, the remaining emoji changes include gender-neutral ballet dancers in different skin tones, as well as more skin tone variants for existing codes like the "people with bunny ears" and "people wrestling" emojis.

Plus: Meta has been accused of burying research in VR's dangers In May, MIT Technology Review published an unprecedented and comprehensive look at how much energy the AI industry uses--down to a single query. Our reporters and editors traced where AI's carbon footprint stands now, and where it's headed, as AI barrels towards billions of daily users. We've just produced a short video to accompany that investigation. You can read the original full story here, and check out--and share-- the full video on YouTube here . AI is changing the grid. Could it help more than it harms?

Video-based spatial cognition is vital for robotics and embodied AI but challenges current Vision-Language Models (VLMs). This paper makes two key contributions. First, we introduce ViCA (Visuospatial Cognitive Assistant)-322K, a diverse dataset of 322,003 QA pairs from real-world indoor videos (ARKitScenes, ScanNet, ScanNet++), offering supervision for 3D metadata-grounded queries and video-based complex reasoning. Second, we develop ViCA-7B, fine-tuned on ViCA-322K, which achieves new state-of-the-art on all eight VSI-Bench tasks, outperforming existing models, including larger ones (e.g., +26.1 on Absolute Distance). For interpretability, we present ViCA-Thinking-2.68K, a dataset with explicit reasoning chains, and fine-tune ViCA-7B to create ViCA-7B-Thinking, a model that articulates its spatial reasoning. Our work highlights the importance of targeted data and suggests paths for improved temporal-spatial modeling. We release all resources to foster research in robust visuospatial intelligence.

Deep neural networks have transformed machine learning, excelling in tasks such as image classification and natural language processing through hierarchical feature learning [1]. However, traditional neurons, which compute deterministic weighted sums followed by nonlinear activations (e.g., ReLU, sigmoid), struggle to model uncertainty or multimodal distributions prevalent in real-world data. This limitation has historically been addressed by probabilistic graphical models, such as Bayesian Networks [2] and Markov Random Fields [3], which offer robust frameworks for uncertainty quantification and complex dependency modeling [4]. These models provide a strong conceptual foundation, but often lack the deep hierarchical feature learning capabilities of modern neural networks. A key research focus has therefore been on bridging the gap between these two paradigms. This has led to approaches that incorporate the probabilistic principles of graphical models directly into deep learning architectures. For example, Bayesian Neural Networks (BNNs) embed uncertainty into network weights [5], while Probabilistic Circuits (PCs), including Sum-Product Networks (SPNs) [7, 8, 9], are deep probabilistic models that build on a formal probabilistic structure, fusing the representational power of graphical models with the hierarchical feature learning of neural networks. In contrast, this paper introduces a novel approach by embedding a univariate Gaussian Mixture Model (uGMM) directly into the network's computational units, enabling each neuron to represent 1

Nested logit (NL) has been commonly used for discrete choice analysis, including a wide range of applications such as travel mode choice, automobile ownership, or location decisions. However, the classical NL models are restricted by their limited representation capability and handcrafted utility specification. While researchers introduced deep neural networks (DNNs) to tackle such challenges, the existing DNNs cannot explicitly capture inter-alternative correlations in the discrete choice context. To address the challenges, this study proposes a novel concept - alternative graph - to represent the relationships among travel mode alternatives. Using a nested alternative graph, this study further designs a nested-utility graph neural network (NestGNN) as a generalization of the classical NL model in the neural network family. Theoretically, NestGNNs generalize the classical NL models and existing DNNs in terms of model representation, while retaining the crucial two-layer substitution patterns of the NL models: proportional substitution within a nest but non-proportional substitution beyond a nest. Empirically, we find that the NestGNNs significantly outperform the benchmark models, particularly the corresponding NL models by 9.2\%. As shown by elasticity tables and substitution visualization, NestGNNs retain the two-layer substitution patterns as the NL model, and yet presents more flexibility in its model design space. Overall, our study demonstrates the power of NestGNN in prediction, interpretation, and its flexibility of generalizing the classical NL model for analyzing travel mode choice.

Survival analysis is a fundamental tool for modeling time-to-event outcomes in healthcare. Recent advances have introduced flexible neural network approaches for improved predictive performance. However, most of these models do not provide interpretable insights into the association between exposures and the modeled outcomes, a critical requirement for decision-making in clinical practice. To address this limitation, we propose Additive Deep Hazard Analysis Mixtures (ADHAM), an interpretable additive survival model. ADHAM assumes a conditional latent structure that defines subgroups, each characterized by a combination of covariate-specific hazard functions. To select the number of subgroups, we introduce a post-training refinement that reduces the number of equivalent latent subgroups by merging similar groups. We perform comprehensive studies to demonstrate ADHAM's interpretability at the population, subgroup, and individual levels. Extensive experiments on real-world datasets show that ADHAM provides novel insights into the association between exposures and outcomes. Further, ADHAM remains on par with existing state-of-the-art survival baselines in terms of predictive performance, offering a scalable and interpretable approach to time-to-event prediction in healthcare.

We investigate reinforcement learning in the Game Of Hidden Rules (GOHR) environment, a complex puzzle in which an agent must infer and execute hidden rules to clear a 6$\times$6 board by placing game pieces into buckets. We explore two state representation strategies, namely Feature-Centric (FC) and Object-Centric (OC), and employ a Transformer-based Advantage Actor-Critic (A2C) algorithm for training. The agent has access only to partial observations and must simultaneously infer the governing rule and learn the optimal policy through experience. We evaluate our models across multiple rule-based and trial-list-based experimental setups, analyzing transfer effects and the impact of representation on learning efficiency.

Existing code large language models (LLMs) often rely on large-scale instruction data distilled from proprietary LLMs for fine-tuning, which typically incurs high costs. In this paper, we explore the potential of small-scale open-source LLMs (e.g., 7B) as synthesizers for high-quality code instruction data construction. We first observe that the data synthesis capability of small-scale LLMs can be enhanced by training on a few superior data synthesis samples from proprietary LLMs. Building on this, we propose a novel iterative self-distillation approach to bootstrap small-scale LLMs, transforming them into powerful synthesizers that reduce reliance on proprietary LLMs and minimize costs. Concretely, in each iteration, to obtain diverse and high-quality self-distilled data, we design multi-checkpoint sampling and multi-aspect scoring strategies for initial data selection. Furthermore, to identify the most influential samples, we introduce a gradient-based influence estimation method for final data filtering. Based on the code instruction datasets from the small-scale synthesizers, we develop SCoder, a family of code generation models fine-tuned from DeepSeek-Coder. SCoder models achieve state-of-the-art code generation capabilities, demonstrating the effectiveness of our method.

Learning rate warmup is a popular and practical technique in training large-scale deep neural networks. Despite the huge success in practice, the theoretical advantages of this strategy of gradually increasing the learning rate at the beginning of the training process have not been fully understood. To resolve this gap between theory and practice, we first propose a novel family of generalized smoothness assumptions, and validate its applicability both theoretically and empirically. Under the novel smoothness assumption, we study the convergence properties of gradient descent (GD) in both deterministic and stochastic settings. It is shown that learning rate warmup consistently accelerates GD, and GD with warmup can converge at most $Θ(T)$ times faster than with a non-increasing learning rate schedule in some specific cases, providing insights into the benefits of this strategy from an optimization theory perspective.