- 资讯公开站Digitimes
Z.ai GLM-5.3 security draws scrutiny as Anthropic flags weak defenses
Anthropic has warned that GLM-5.3, the latest model from Chinese AI startup Z.ai, approached Claude Mythos Preview on an exploit-development benchmark but lacks sufficient safety protections, raising concerns that malici…
- 资讯公开站Digitimes
South Korea's AI chip budget goes unspent through August
South Korea's government is pushing to join the "global top three AI powers," but the AI semiconductor development program led by the Ministry of Trade, Industry and Resources (MOTIR) had spent nothing by the end of Augu…
- 资讯公开站Digitimes
Weekly news roundup: Musk looms large as AI chip boom drives capacity expansion and geopolitical risks
Across the week, Taiwan-linked semiconductor groups and their partners continued expanding capacity, while AI demand, export controls, and financing structures shaped the industry outlook. ASML's push to raise EUV output…
- 资讯公开站Digitimes
Taiwanese aerospace company AIDC unveils Raven X tactical drone
Aerospace Industrial Development (AIDC) unveiled its first self-developed fixed-wing tactical drone, the Raven X. AIDC president Hsiu-mei Chuang recently stated that the Taiwan-based aerospace company is shifting part of…
- 资讯公开站Digitimes
Japan to finalize AI growth investment plan this year
Japan plans to detail the first five years of Prime Minister Sanae Takaichi's more than JPY370 trillion (US$2.3 trillion) 14-year investment initiative, which is designed to spur investment in AI and other strategic sect…
- 论文公开站arXiv
Revisiting Input Time-frequency Representations in Multi-pitch Estimation for Vocal Ensembles
Multi-pitch estimation in vocal ensembles is challenging because singers occupy overlapping pitch ranges and often sing at closely spaced fundamental frequencies, causing their harmonics to overlap in time-frequency repr…
- 论文公开站arXiv
Forecasting from Counterfactual Simulator Rollouts: A Sim2Real Evaluation
Deploying a new decision policy creates a cold-start problem for prediction models whose targets depend on the policy's actions: historical observations reflect earlier policies, while real observations under the new pol…
- 论文公开站arXiv
Pivot-SD: Efficient Self-Distillation for Masked Diffusion Language Models
Masked diffusion language models (dLMs) offer a promising parallel alternative to autoregressive models for complex reasoning. However, they face a distinct credit-assignment challenge, since a few commitments during den…
- 论文公开站arXiv
Planning to Learn
Policy-gradient methods are central to modern reinforcement learning, including LLM post-training. When they struggle, the usual suspects are exploration, credit assignment and action-sampling noise. Classification has n…
- 论文公开站arXiv
FrugalEvo: Towards Cost-Aware LLM-Guided Program Evolution
LLM-guided evolutionary methods, such as AlphaEvolve, have emerged as powerful approaches for challenging computational optimization problems, such as circle packing. However, prior work typically optimizes performance g…
- 论文公开站arXiv
Simulation-Free Learning of Population Dynamics with Wasserstein Lagrangian Residuals
The dynamics of cells, organisms, and fluids are often modeled as probability distributions evolving over time. Reconstructing and extrapolating this evolution from unpaired snapshots requires assumptions about the under…
- 论文公开站arXiv
Risk-Aware Input-Constrained Safe Intercept Guidance Against Multiple Moving Defenders
This paper develops a risk-aware guidance law for an attacker to intercept a stationary target in the presence of multiple moving defenders while respecting the control input limits. We represent defender threats via eng…
- 论文公开站arXiv
Transcriptome-informed multi-modal AI for predicting neoadjuvant therapy response from breast cancer biopsies
Scarcity of labeled data limits development of deep learning biomarkers in oncology. We develop a two-stage AI model predicting pathological complete response (pCR) to neoadjuvant therapy in breast cancer. The first stag…
- 论文公开站arXiv
LESSER: Post-Training Data Selection with Output-Layer Gradients
The choice of post-training data for large language models substantially affects downstream performance. Gradient-based data selection is a popular approach that ranks training data by how well their gradients align with…
- 论文公开站arXiv
From Mixing to Tearing: Graph Decomposition in Decentralized Optimization via Message Passing
We study the minimization of sums of smooth strongly convex functions over undirected graphs, with each function held by one agent and communication restricted to neighbors in the graph. Existing decentralized methods, w…
- 论文公开站arXiv
EyeRobot 2.0: Active Gaze for Precise Manipulation without Wrist Cameras
Inspired by human vision, we introduce a framework using active gaze to enable fine-grained bimanual manipulation with only a single stereo camera. EyeRobot 2.0 physically attends to a 3D fixation point in the scene by s…
- 论文公开站arXiv
RNADyn: A Benchmark for Generating and Understanding RNA Dynamics
Ribonucleic acid (RNA) functions through conformational changes that are not fully captured by static structures. However, large-scale standardized RNA dynamics data remain limited, and existing approaches typically trea…
- 论文公开站arXiv
What Should World Models Forget? Stratified Retention for Continual Adaptation
Continual learning treats degradation on previously seen data as evidence of failure, a convention inherited from settings with a stationary prediction target, where a correct label remains correct indefinitely. World mo…
- 论文公开站arXiv
4DCodeBench: Benchmarking Agents on Inverse Graphics of Dynamic Scenes
We introduce 4DCodeBench, a benchmark for 4D inverse graphics through code generation, in which agents reconstruct dynamic scenes from video as executable graphics programs. To accomplish this, agents must translate visu…
- 论文公开站arXiv
Less Decoder is More Encoder: Geometric Representation Learning from Novel View Synthesis
This paper examines the role of Novel View Synthesis (NVS) in geometric representation learning. In principle, NVS should reason about 3D scene structure, thereby enabling transferable multi-view geometric representation…
- 资讯公开站Entrepreneur
Better Branding Will Make More Money Than Having the Best Marketing — Lessons Learned From Real Estate
As consumers become overwhelmed with options, building trust before the first sale will become the obvious choice when it matters most.
- 资讯公开站CNX Software
Geekworm UPS Scap 5V5A – A supercapacitor UPS HAT for Raspberry Pi 5
Geekworm, in collaboration with Mcuzone, has launched the UPS Scap 5V5A, a new supercapacitor UPS HAT for the Raspberry Pi 5 designed to keep the Raspberry Pi running long enough (a few seconds to a few minutes) to save …
- 资讯公开站Digitimes
SpaceX's Starship reaches orbit, but engine issues raise stakes for NASA's Moon plan
SpaceX's Starship reached Earth orbit for the first time on its 14th test flight and successfully deployed 26 third-generation Starlink satellites, marking a major milestone for the heavy-lift rocket. Engine anomalies du…
- 资讯公开站Digitimes
DIGITIMES Insight: AI data demand surges as power and cooling become new bottlenecks
As the number of AI computing transistors integrated within a single CoWoS package rises rapidly, DIGITIMES observes that moving massive volumes of data to compute chips fast enough is becoming increasingly critical to f…
- 资讯公开站Digitimes
Tesla leads Taiwan registrations as it advances FSD road tests
Tesla ranked first in Taiwan's new-car registrations in September and said it is continuing the approval process for its Full Self-Driving (Supervised), or FSD (S), driver-assistance system. The automaker also presented …
