
Nvidia says its open-source AI strategy remains intact after Jensen Huang’s exit
Nvidia's interim CEO reaffirmed its open-source AI strategy and plans to acquire Hugging Face, despite Jensen Huang's sudden departure.

Nvidia's interim CEO reaffirmed its open-source AI strategy and plans to acquire Hugging Face, despite Jensen Huang's sudden departure.

Despite the strong growth, Vietnam still accounts for only a small share of Russia’s total banana imports, which amounted to 1.3 million tonnes in the first months of 2026. Ecuador remains the dominant supplier, with more than 92% of the market, or 1.2 million tonnes, up 8.3% year on year. Ecuadorian bananas also made up over 95% of supplies to major Russian retail chains. Vietnam step up state audit cooperation with Russia, Belarus New direct air link strengthens…

Nvidia posted a stronger-than-expected third quarter and a bullish outlook in its first earnings report after Jensen Huang was ousted as CEO in a scandal that has reportedly spawned an SEC investigation. Why it matters: Nvidia remains squarely at the center of the AI economy — and investors are anxious for guidance on the company's strategic roadmap following Huang's sudden exit. Zoom in: Nvidia recorded quarterly revenue of $115.8 billion and net income of $65.3 billion…

Hisense's UR8 brings RGB Mini LED technology to a more accessible price point, with excellent brightness and impressive blooming control. However, its real-world color performance leaves room for improvement, while OLED remains a compelling alternative.
Time-series sensing increasingly relies on neural networks, yet deployment on resource-constrained microcontrollers remains challenging. Aggressive model compression can reduce accuracy, while restricting weights to multiplication-free representations introduces further losses. We present BeFit, an end-to-end framework that shifts model capacity from weights to learnable Bernstein polynomial activations. BeFit (i) finds compact Bernstein-activated convolutional neural…
YouTube Music is simplifying its core navigation on Android and iOS by removing the Samples tab, though the vertical video format remains in the app.
For most of its history, the high-performance computing market measured progress the same way: how fast a machine could solve a fixed problem. Rankings rose, a new system took the top spot, and the cycle started over. Raw performance remains the foundation of scientific computing, and Nvidia Corp. brought plenty of it to SC26. But

The remains of a 37-year-old woman who vanished more than two years ago in the New Jersey Pine Barrens were found Saturday during a multiagency search, state police said.
Robot manipulation videos provide a rich visual experience, yet recovering robot action and gripper-state annotations from RGB observations still remains challenging. Using the embodiment-agnostic tool-center-point (TCP) pose as the action space, we present TCP4D, a unified model in which a shared 4D scene representation both tracks observed TCP motion and generates future TCP trajectories. Given a monocular video and initial 2D TCP queries, the tracker directly recovers…
The identifications are part of an ongoing effort to locate and identify the remains of U.S. soldiers killed in the Korean War.

Researchers at Korea Advanced Institute of Science and Technology (KAIST) published “Clock Distribution Network with Backside Clock Mesh in BSPDN Architecture.” Abstract Excerpt “We propose a clock distribution network (CDN) in which the clock mesh is located on backside metal (BM) layers, while the rest of the network remains on the frontside. Since the backside... »
Hierarchical models have been proposed as synthetic settings for studying language. However, their relevance to realistic language corpora remains underexplored. In this work, we introduce tree-grams, autoregressive generative $n$-gram models that build the conditional probability tensor $p(x_n \mid x_{1:n-1})$ through a hierarchical composition of order-2 and order-3 tensors. $n$ controls the maximal context window, whereas a rank hyperparameter $d$ controls model capacity…
AI-assisted RTL localization and debugging need ground-truth bug datasets with known locations and observable failures, while testbench grading uses injected bugs to measure detection capability. Creating such datasets at scale remains a major EDA challenge. BugGen 1.0 shows that Large Language Models (LLMs) can generate context-aware RTL mutations. Still, it relies on the model for both choosing where to mutate and generating the edit, which increases inference costs and…
This research-to-practice study examines the use of student-created instructional videos as an active learning strategy in a large undergraduate Software Design and Engineering (SWE) course. Supporting engagement and understanding in large-scale computer science (CS) courses remains a challenge, as instruction often relies on automated activities and summative assessments that prioritize grading over meaningful learning. This can make the learning experience feel impersonal…

Amid the war in Iran, Saudi Arabia's human rights situation remains bleak, with at least 211 executions, corporal punishment and an intensifying crackdown on critics. Is the regional conflict driving domestic repression?

A new House majority can push climate policy, but a full US return remains uncertain.
Vulnerability exploitation tests whether language model agents can develop working exploits from known software flaws. Sequential test-time scaling can improve exploitation performance, but larger budgets do not ensure sustained exploration: agents often stop with budget remaining. Yet what happens when execution continues beyond stopping remains underexplored. We investigate this gap on ExploitBench using DeepSeek-V4-Flash-0731 and MiMo-V2.6-Pro. We find additional verified…

Vehicles weighing up to 10 tonnes and public transport up to 18 tonnes can use the bridge towards the right bank. Traffic in the opposite direction remains closed.

Local authorities in the Russian city of Kursk have installed a sculpture dedicated to DPRK military personnel participating in the war against Ukraine, though it currently remains under protective film.
For automatic optimization systems, performance portability across increasingly diverse AI accelerators remains a challenge. We investigate how well graph-level superoptimization transfers across architectures by adapting Mirage from its CUDA-oriented design to the Huawei Ascend 910B3 through a Triton-based compilation stack. We evaluate the adapted system on transformer-relevant workloads, including matrix multiplication, softmax, RMSNorm+MatMul, gated MLP, FNN, and…

Tehran, IRNA – Foreign Minister Abbas Araghchi has emphasized that Iran has always taken diplomacy seriously and has remained firm in defending its territorial integrity and national interests.
Our aim is to accelerate autoregressive vision-language-action models (VLAs), which generate robot actions one token at a time. Speculative decoding makes this possible: a lightweight draft model proposes several tokens at once, and the VLA verifies them in a single pass, accepting only those that match its own predictions, so decoding remains lossless. The speedup, however, is limited in robot control, where many different tokens encode nearly the same motion and a draft…
GitHub natively supports workflow automation through GitHub Actions, now the dominant CI/CD tool on the platform. Yet keeping these workflows secure remains an ongoing burden: maintainers often patch a vulnerability on the default branch but leave the same flaw on the release and maintenance branches that downstream consumers continue to build from. Carrying a known fix onto these divergent branches, called backporting, is rarely done, and this paper presents the first…
Reliable localization in long underground infrastructures remains a major challenge for autonomous inspection robots. In hydropower spillway tunnels and drainage galleries, repetitive geometric structures provide insufficient constraints along the tunnel axis, while weak texture and illumination variations further degrade visual estimation. Under this coupled degradation, existing LiDAR-inertial and visual-inertial odometry systems may lose the ability to recover continuous…
Robust watermarking is commonly verified by asking whether a hidden payload remains decodable after compression, geometric distortion, or generative editing. Localized and semantic extensions may additionally check whether an expected region or category is still present. These tests establish evidence validity or content consistency, but not whether the evidence still belongs to the particular visual instance registered by its owner. We expose cross-instance transplantation…
The group is expanding production capacity and investing in technology to strengthen the competitiveness of Vietnamese manufacturers.
Analog circuit design automation requires both coordinated tool execution and efficient numerical search. Large language models (LLMs) offer new ways to organize tool execution, but finding one sizing that satisfies specifications across process, voltage, and temperature (PVT) corners remains costly. PVTSizing 2.0 addresses these needs through a human-agent workflow built around an independently usable numerical optimizer. An agent carries the designer's approved task…

The remains of a top Australian racehorse that died of colic on Monday were found at a public rubbish tip.
Multimodal large language models (MLLMs) are rapidly advancing, yet their safety remains a critical concern, especially under jailbreak attempts. Existing MLLM jailbreaks either inject fragile adversarial perturbations into irrelevant images or synthesize toxic visual cues to achieve text-driven attack intents. In contrast, a more practical MLLM jailbreak emphasizes enhancing attack efficacy directly on naturally occurring harmful image-text pairs. We address this problem by…
In traditional recommender systems (RSs), textual information typically serves as auxiliary signals for representation learning. In contrast, in LLM-based RSs, textual information becomes the primary medium for recommendation generation through semantic understanding and reasoning. This shift fundamentally changes the attack surface, making promotion-driven manipulations more feasible, effective, and harder to detect. However, effectively detecting such attacks remains…