https://www.intlnews.top/technology/ Global News Digest - Latest Technology News Tue, 02 Jun 2026 06:50:15 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_Schroders-Renewable-Unit-Targets-AI-Assets-as-Powe_418/ Schroders’ renewable unit targets AI infrastructure, pivoting to meet soaring energy demand from artificial intelligence assets. Tue, 02 Jun 2026 03:54:51 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_State-Streets-Paglia-on-SBI-Group-Partnership-ETFs_273/ State Street's Paglia discusses the SBI Group partnership and ETFs, but the source text is missing. Please provide the article to generate a summary. Tue, 02 Jun 2026 03:37:33 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_Nvidia-Boss-Says-Workers-Should-Be-Paid-as-Much-as_427/ Nvidia CEO Jensen Huang advocates for paying workers “as much as possible,” emphasizing maximum compensation. This stance highlights his commitment to employee welfare amidst the company's significant industry influence and success. Tue, 02 Jun 2026 03:33:16 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_TSE-Talking-With-Regulator-For-Easing-ETF-Listing-_269/ The Tokyo Stock Exchange is discussing with regulators to ease ETF listing rules. This aims to simplify market access and boost liquidity for exchange-traded funds. Tue, 02 Jun 2026 03:32:47 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_SP-DJI-CEO-on-Japan-Markets-Mega-IPOs_271/ S&P DJI CEO discusses Japan's financial markets and major IPOs. Tue, 02 Jun 2026 03:28:46 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_ABB-Robotics-on-Nvidia-Partnership-Physical-AI_272/ ABB Robotics discusses its collaboration with Nvidia, focusing on advancing Physical AI capabilities to enhance robotic automation and intelligence. Tue, 02 Jun 2026 03:26:37 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_China-Adds-Data-and-AI-to-Trade-Secret-Rules-to-Bl_274/ China has expanded trade secret regulations to include data and AI, aiming to prevent leaks and protect intellectual property. Tue, 02 Jun 2026 02:46:46 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_Nvidia-Chips-Sought-by-Chinese-Labs-With-Military-_276/ Chinese labs with military ties are seeking Nvidia chips, raising concerns about potential dual-use technology transfers. This pursuit highlights ongoing tensions over semiconductor exports to entities linked to China's defense sector. Tue, 02 Jun 2026 02:19:04 +0000 https://www.intlnews.top/technology/2026/06/02/bloomberg_Chinese-Bionic-Hand-Maker-Sees-Humanoid-Robot-Led-_275/ BrainCo reports a sales surge for its bionic hands, driven by demand from humanoid robots. This boom highlights the growing commercial viability of advanced prosthetic technology in the robotics sector. Tue, 02 Jun 2026 02:01:50 +0000 https://www.intlnews.top/technology/2026/06/02/reuters_Asia-stocks-skittish-as-Middle-East-anxiety-offset_142/ Asian stocks faced volatility as Middle East tensions offset AI-driven optimism, creating a mixed market sentiment. Tue, 02 Jun 2026 00:38:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_When-AI-Benchmarks-Plateau-A-Systematic-Study-of-B_1950/ This study finds that nearly half of AI benchmarks have saturated, correlating with age. Expert curation, not public data, drives resilience, guiding robust evaluation design. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Acting-with-AI-An-Interaction-Based-Framework-for-_1361/ This paper proposes an interaction-based framework for AI tort liability, categorizing harm into autonomous drift, tool use, and collaborative planning. It uses interaction logs to assign responsibility, resolving a key gap in current tort law. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_TAPS-Target-Aware-Prefix-Tree-Selection-for-Diffus_1360/ TAPS optimizes speculative decoding by selecting target-aware prefix trees, reducing verification latency. It achieves up to 7.9x speedup over autoregressive decoding, outperforming state-of-the-art methods. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_KACE-Knowledge-Adaptive-Context-Engineering-for-Ma_1359/ KACE decouples knowledge storage from usage via an epistemic tree, dynamically retrieving relevant guidance to solve math problems. It achieves 62.2% accuracy on AIME 2025, outperforming Best-of-N methods with similar computational costs. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Threshold-Based-Exclusive-Batching-for-LLM-Inferen_1358/ This paper introduces EB+, a hybrid scheduler that dynamically switches between exclusive and mixed batching based on hardware bandwidth. EB+ outperforms static strategies, achieving up to 36.4% higher throughput across varying workloads and hardware configurations. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_EnergyMamba-An-Uncertainty-Aware-Graph-Enhanced-Se_1357/ EnergyMamba integrates graph-enhanced selective state space models with adaptive quantile regression to capture spatiotemporal correlations and uncertainty. It outperforms 15 baselines, improving accuracy by 5% and uncertainty quantification by 6% across real-world datasets. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_VESTA-Visual-Exploration-with-Statistical-Tool-Age_1356/ VESTA enhances statistical modeling by empowering VLMs with dynamic, self-generated diagnostic tools. It outperforms existing methods on complex tasks, validated via the new DAWN benchmark. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_SDR-Set-Distance-Rewards-for-Radiology-Report-Gene_1355/ SDR uses set-distance rewards to improve radiology report generation, outperforming SFT and exact-match methods. It also enables efficient test-time scaling via candidate pruning. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Doing-What-They-Say-Not-What-They-Reason-Locating-_1354/ This study uses a Texas Hold’em simulator to quantify the faithfulness gap in LLM agents, revealing opposing behaviors in reasoning-to-conclusion versus conclusion-to-action stages. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Weak-Critics-Make-Strong-Learners-On-Policy-Critiq_1353/ This paper introduces On-Policy Critique Distillation (OPCD), using weak models as critics to guide strong LLMs. This "weak-critic strong oversight" approach boosts performance by filtering high-quality critiques and transferring behaviors via adaptive self-teaching. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Coupling-Language-Models-with-Physics-based-Simula_1352/ This study integrates LLMs with physics-based simulations to plan inorganic material synthesis. Results show LLMs generate more feasible strategies than traditional algorithms, leveraging their implicit knowledge. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_The-Deterministic-Horizon-When-Extended-Reasoning-_1351/ Extended reasoning fails deterministic tasks due to architectural limits, creating a "Deterministic Horizon" where tool delegation is essential. Hybrid approaches significantly outperform pure neural methods, confirming an inherent capability ceiling. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_From-Noise-to-Control-Parameterized-Diffusion-Poli_1350/ Parameterized Diffusion Policy (PDP) conditions diffusion on a learned behavior manifold, enabling precise control and smooth interpolation. It outperforms standard policies by synthesizing novel behaviors without weight updates. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_From-Weak-Signals-to-Strong-Models-Preference-Delt_1349/ Preference Delta Aggregation merges LoRA adapters from weak model pairs to boost strong LLMs. Combined with Geometric Alignment Merging, it significantly outperforms baselines in reasoning and agentic search. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Model-Native-Computing-Architecture-Envisioning-Fu_1348/ This paper proposes ICAM, a six-layer framework for model-native computing, resolving LLM roles via a dual-plane perspective. It introduces three design laws to address cache, context, and agent efficiency challenges in future system architectures. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Evaluating-Bivariate-Causal-Statements-Based-on-Mu_1347/ This study proposes compatibility and incompatibility scores to evaluate bivariate causal statements without relying on ground truth. These metrics effectively distinguish accurate from erroneous claims, aiding validation of insights from humans or AI. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_On-Wednesdays-We-Ask-Questions-Optimizing-Active-L_1346/ The study finds that while cheap LLMs classify well, high-cost models like GPT-5 are needed for effective legal triage questions. However, inconsistencies in specific areas like domestic violence highlight the need for specialized screening modules. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Robust-Shielding-for-Safe-Reinforcement-Learning_1345/ This paper introduces robust shielding for safe reinforcement learning in unknown environments. It guarantees safety under worst-case transitions while allowing optimal agent behavior. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_MindZero-Learning-Online-Mental-Reasoning-With-Zer_1344/ MindZero enables efficient online mental reasoning in MLLMs via self-supervised reinforcement learning, eliminating annotation needs. It outperforms model-based methods in speed and accuracy for real-time AI assistance. Tue, 02 Jun 2026 00:00:00 +0000 https://www.intlnews.top/technology/2026/06/02/arxiv_Capability-Self-Assessment-Teaching-LLMs-to-Know-T_1343/ This study shows LLMs overestimate competence, but reinforcement learning effectively teaches Capability Self-Assessment (CSA) without degrading performance. CSA generalizes well and improves decision-making and training data selection. Tue, 02 Jun 2026 00:00:00 +0000