Summer Game Fest just wrapped up its sixth year and, like a beautifully cel-shaded version of The Blob, the show just keeps on growing. The official Summer Game Fest 2026 showcase took place on June 5, but the surrounding buffet of new game reveals, release date announcements, review opportunities and developer spotlights actually ran from June 1 all the way to June 9. That's more than an entire week of near-constant video game news and trailers to consume, and here we've gathered absolutely all of it in one tidy but lengthy package. First up, a collection of Engadget's previews and reporting from Summer Game Fest Play Days in Los Angeles, which ran from June 6-8: Control Resonant's take on New York feels like the Backrooms Silent Hill Townfall brings atmospheric horror to '90s Scotland with incredible attention to detail Saw: Genesis looks the most fun when you're the murderous mastermind Alien: Isolation 2 keeps the classic horror game's uncompromising approach to raising tension Spyro: A Realm Beyond sees the '90s purple dragon make a big comeback Be like Carl from Summer House and get in the MIX with another high-energy stream filled with great-looking upcoming indie games, gathered by the folks at the Media Indie Exchange. The MIX hosts a smattering of annual online indie showcases, and alongside in-person events, they've been spreading the good gaming word for the past 10 years.

The effectiveness of domain translation in addressing image-based problems of Unsupervised Domain Adaptation (UDA) depends on the quality of the translated images and the preservation of crucial discriminative features. However, achieving high-quality and stable translations typically requires paired data, which poses a challenge in scenarios with limited annotations in the target domain. To address this issue, this paper proposes a novel method termed Stain-Guided Cycle Diffusion (SGCD), employing a dual diffusion model with bidirectional generative constraints to synthesize highly realistic data for downstream task fine-tuning. The bidirectional generative constraints ensure that the translated images retain the features critical to the downstream model in properly controlling the generation process. Additionally, a stain-guided consistency loss is introduced to enhance the denoising capability of the dual diffusion model, thereby improving the quality of images translated between different domains using latents from one domain and a diffusion model trained on another. Experiments conducted on four public datasets demonstrate that SGCD can effectively enhance the performance of downstream task models on the target domain.

Crystal structures are defined by the periodic arrangement of atoms in 3D space, inherently making them equivariant to SO(3) group. A fundamental requirement for crystal property prediction is that the model's output should remain invariant to arbitrary rotational transformations of the input structure. One promising strategy to achieve this invariance is to align the given crystal structure into a canonical orientation with appropriately computed rotations, or called frames. However, existing work either only considers a global frame or solely relies on more advanced local frames based on atoms' local structure. A global frame is too coarse to capture the local structure heterogeneity of the crystal, while local frames may inadvertently disrupt crystal symmetry, limiting their expressivity. In this work, we revisit the frame design problem for crystalline materials and propose a novel approach to construct expressive {\bf S}ymmetry Preserving Frames, dubbed as SPFrame, for modeling crystal structures.

Online reinforcement learning (RL) excels in complex, safety-critical domains but suffers from sample inefficiency, training instability, and limited interpretability. Data attribution provides a principled way to trace model behavior back to training samples, yet existing methods assume fixed datasets, which is violated in online RL where each experience both updates the policy and shapes future data collection. In this paper, we initiate the study of data attribution for online RL, focusing on the widely used Proximal Policy Optimization (PPO) algorithm. We start by establishing a attribution framework, interpreting model checkpoints with respect to the records in the recent training buffer. We design two target functions, capturing agent action and cumulative return respectively, and measure each record's contribution through gradient similarity between its training loss and these targets. We demonstrate the power of this framework through three concrete applications: diagnosis of learning, temporal analysis of behavior formation, and targeted intervention during training. Leveraging this framework, we further propose an algorithm, iterative influence-based filtering (IIF), for online RL training that iteratively performs experience filtering to refine policy updates. Across standard RL benchmarks (classic control, navigation, locomotion) to RLHF for large language models, IIF reduces sample complexity, speeds up training, and achieves higher returns. Together, these results open a new direction for making online RL more interpretable, efficient, and effective.

Recent models such as OpenAI o1 and DeepSeek-R1 have demonstrated strong performance on reasoning-intensive tasks by generating extended Chain-of-Thought (CoT) traces. While longer reasoning helps with thorough exploration of solution paths for complex problems, it also often leads to inefficient and redundant outputs--a phenomenon commonly described as $\textit{overthinking}$. In this paper, we propose $\texttt{ShorterBetter}$, a simple yet effective reinforcement learning method that enables reasoning models to learn their own optimal CoT lengths without manual supervision. We define the $\textit{Sample Optimal Length}$ (SOL) as the length of the shortest correct response among multiple generations, which serves as a dynamic reward signal to guide the model toward efficient reasoning. Applied to DeepSeek-Distill-Qwen-1.5B/7B as base models, $\texttt{ShorterBetter}$ achieves 50\%-80\% reduction in output lengths in both in-domain and out-of-domain reasoning tasks while maintaining accuracy. Our reasoning trace analysis shows that $\texttt{ShorterBetter}$ refines the structure of the reasoning traces by reducing unnecessary repetition, excessive self-verification, and over-exploration of alternatives.

Microsoft released PowerToys version 0.100, celebrating the utility suite's 20th anniversary with significant feature improvements and performance enhancements. PCWorld reports the update introduces a dynamic Shortcut Guide, enhanced Command Palette, and Extension Gallery for community-developed tools. Key improvements include faster startup times, better multi-monitor support, improved Performance Monitor, and a 15% smaller installer through .NET 10 upgrade. Marking the 20th anniversary of PowerToys, Microsoft just released version 0.100 of this extremely useful collection of Windows tools and features. If you aren't familiar with PowerToys yet, check out our picks for the most useful PowerToys features you might've missed .

Recently proposed Fine-Grained Few-Shot Class-Incremental Learning (FG-FSCIL) offers a practical and efficient solution for enabling models to incrementally learn new fine-grained categories under limited data conditions. However, existing methods still settle for the fine-grained feature extraction capabilities learned from the base classes.

We study nonlinearly preconditioned gradient methods for smooth nonconvex optimization problems, focusing on sigmoid preconditioners that inherently perform a form of gradient clipping akin to the widely used gradient clipping technique. Building upon this idea, we introduce a novel heavy ball-type algorithm and provide convergence guarantees under a generalized smoothness condition that is less restrictive than traditional Lipschitz smoothness, thus covering a broader class of functions. Additionally, we develop a stochastic variant of the base method and study its convergence properties under different noise assumptions. We compare the proposed algorithms with baseline methods on diverse tasks from machine learning including neural network training.

When you purchase through links in our articles, we may earn a small commission. The Baseus Picogo AC22 is the size of an earbud case, charges at 45W, and has a built-in cable. The Baseus Picogo AC22 mini power bank is easily one of the tiniest models you'll find, making it a top choice for traveling and everyday use. Again, it's the size of this power bank that makes it stand out. It measures 2.8 x 2.4 x 1.0 inches--equivalent to a wireless earbuds case--and it only weighs 0.375 pounds.

Memory Mosaics, networks of associative memories, have demonstrated appealing compositional and in-context learning capabilities on medium-scale networks (GPT-2 scale) and synthetic small datasets. This work shows that these favorable properties remain when we scale memory mosaics to large language model sizes (llama-8B scale) and real-world datasets. To this end, we scale memory mosaics to 10B size, we train them on one trillion tokens, we introduce a couple architectural modifications (), we assess their capabilities across three evaluation dimensions: training-knowledge storage, new-knowledge storage, and in-context learning. Throughout the evaluation, memory mosaics v2 match transformers on the learning of training knowledge (first dimension) and significantly outperforms transformers on carrying out new tasks at inference time (second and third dimensions). These improvements cannot be easily replicated by simply increasing the training data for transformers. A memory mosaics v2 trained on one trillion tokens still perform better on these tasks than a transformer trained on eight trillion tokens.