Vision-Language Models (VLMs) achieve strong performance on spatial question answering benchmarks, yet it remains unclear whether such gains reflect genuine spa...
Reinforcement learning with verifiable rewards has made post-training highly effective when correctness can be checked automatically. However, many important mo...
Non-destructive testing of aerospace SiC/SiC composites via X-ray computed tomography (XCT) relies on expert visual assessment, with current workflows offering ...
Large Vision Language Models (LVLMs) show promise in medical applications, but their inability to faithfully ground responses in visual evidence raises serious ...
Music streaming fraud, where bad actors artificially inflate stream counts to manipulate chart rankings and royalty payments, poses a significant threat to stre...
The proliferation of generative artificial intelligence has given rise to an interactive learning environment, where model parameters are continuously updated u...
Training 3D Gaussian Splatting (3DGS) at billion-primitive scale is fundamentally memory-bound: each Gaussian primitive carries a large attribute vector, and th...
Large language models (LLMs) are widely used for open-ended tasks, but underspecified prompts can lead to low-quality answers and additional interaction. This p...
Text-to-Image (T2I) models have recently seen notable progress around 1K and 2K resolution. With the extreme desire for better visual experience and the rapid d...
Bayesian optimization (BO) selects evaluation points for expensive black-box objectives using Gaussian process (GP) predictive distributions. Kernel choice and ...
Learning generalizable trajectory representations from raw GPS traces remains difficult because the data is continuous, noisy, and irregularly sampled. Spatial ...
Distributed acoustic sensing (DAS) systems generate continuous, ultra-high-channel-count data streams at rates that exceed the capabilities of conventional batc...
Large language models (LLMs) are increasingly integrated into high-stakes decision-making. Inspired by the theory of inattentional blindness in human cognition,...
Artificial vision models are often evaluated against the human visual cortex by measuring how accurately their internal representations predict brain responses....
AI-assisted theorem proving can now generate substantial Lean developments for olympiad-level mathematics, but the evidential status of such developments depend...
We show that time series foundation models scale: a single training recipe produces reliable forecast-quality improvements from 4M to 2.5B parameters. We releas...
Referring segmentation grounds natural-language queries to pixel-level masks, but extending it to complex scenarios with multiple instances, cross-category grou...
Conversational AI has now reached billions of users, yet existing datasets capture only what people say, not what they think. We introduce ThoughtTrace, the fir...
Recent work pairs LLMs with evolutionary search to iteratively generate, modify, and select code using task-specific feedback. These systems have produced stron...
Large language models (LLMs) can enhance factuality via retrieval-augmented generation (RAG), but applying RAG to every query is unnecessary when the model-only...
Chain-of-thought (CoT) is a standard approach for eliciting reasoning capabilities from large language models (LLMs). However, the common CoT paradigm treats th...
Explicit software architecture models are essential artifacts for communicating, analyzing, and evolving complex software-intensive systems. In ROS~2-based robo...
As autonomous coding agents see rapid adoption, their evaluation has primarily focused on task completion rates holding the target codebase fixed. This leaves a...
Backpropagation with gradient descent is a common optimization strategy employed by most neural network architectures in machine learning. However, finding opti...
Documentation has long guided computer system tuning by distilling expert knowledge into per-parameter recommendations. Yet such guides capture only what expert...
Mixture-of-Expert (MoE) models enable efficient inference by employing smaller experts and activating only a subset of them per token. MoE serving engines distr...
Photonic neuromorphic computing offers a promising route to overcoming the limitations of conventional von Neumann architectures by exploiting the high bandwidt...
Dramatic cost reductions driven by private sector innovations have led to a rapid increase in the number of satellites in orbit and a corresponding surge in spa...
Network Intrusion Detection System is a critical means of ensuring cybersecurity. However, existing Genetic Algorithm-based feature selection methods face sever...
LLM discovery and optimization systems are increasingly applied across domains, implementing a common propose-evaluate-revise loop. Such optimization or discove...
We present OpenComputer, a verifier-grounded framework for constructing verifiable software worlds for computer-use agents. OpenComputer integrates four compone...
Can a single LLM-based optimization system match specialized tools across fundamentally different domains? We show that when optimization problems are formulate...
Modern deployments of Large Language Models (LLMs) increasingly require serving multiple models with diverse architectures, sizes, and specialization on shared,...
Google I/O 2026 is just a few hours away, with the big keynote scheduled for May 19. Last year’s event saw a heavy focus on AI, and we are expecting a similar t...
Threat Landscape They discuss the new wave of phishing attacks that target identity and use social engineering, how foreign bodies are exploiting security flaw...
The rapid proliferation of AI-generated visual media has created an urgent need for efficient, trustworthy deepfake detection systems. However, existing deep le...
!https://cdn.platum.kr/wp-content/uploads/2026/05/image-77-1009x1024.png 프로젝트 개요 중소벤처기업부는 3월 26일부터 5월 15일까지 ‘모두의 창업 프로젝트’를 진행했으며, 총 62,944명이 신청했다. 이는 일반·기술 트랙 5...
Redis Iris: A Context and Memory Platform for Agentic AI Redis built its name as the caching layer that kept web applications from collapsing under load. The p...
Physical computing systems provide a promising route toward hardware-native machine learning, but their computational capabilities remain difficult to character...
Existing neural combinatorial optimization solvers frame solution search as imitation of optimal decisions, inherently limiting their utility to single-objectiv...
Layoffs and Employee Morale Meta is expected to begin cutting about 8,000 jobs this week as it pours more money into AI infrastructure and looks to “offset” ot...
Elon Musk, the world's richest man, has not been winning in court lately. His loss on Mondayhttps://www.bbc.com/news/articles/cewpyv79pw1o in his lawsuit agains...
Multiview 3D evaluation assumes that the images being scored are observations of one static 3D scene. This assumption can fail in NVS and sparse-view reconstruc...
Current hierarchical attention methods, such as NSA and InfLLMv2, select the top-k relevant key-value (KV) blocks based on coarse attention scores and subsequen...
Pipeline parallelism is a key technique for scaling large-model training, but modern workloads exhibit runtime variability in computation and communication. Exi...
Modern audio generation predominantly relies on latent-space compression, introducing additional complexity and potential information loss. In this work, we cha...
Recent video editing models have converged on a unified conditioning design: a single diffusion transformer jointly consumes text, source video, and reference i...
Recent large language models (LLMs) have demonstrated strong capabilities in understanding and generating code, from competitive programming to repository-level...
