We study dense and mixture-of-experts (MoE) transformers in a tiny-scale pretraining regime under a shared LLaMA-style decoder training recipe. The sparse model...
Introduction Titanic shipwreck was a major historical incident that shaped how we view human survival during disasters. Even a century later, this tragic incid...
When an omnimodal large language model accepts a question whose textual premise contradicts what it actually sees or hears, does the failure lie in perception o...
LLMs are widely adopted in production, pushing inference systems to their limits. Disaggregated LLM serving (e.g., PD separation and KV state disaggregation) im...
EEG foundation models achieve state-of-the-art clinical performance, yet the internal computations driving their predictions remain opaque: a barrier to clinica...
Large Language Models (LLMs) are widely applied in educational practices, such as for generating children's stories. However, the generated stories are often to...
Federated learning enables collaborative model training across distributed clients, yet vanilla FL exposes client updates to the central server. Secure-aggregat...
LLM-as-a-judge is now the default measurement instrument for open-ended generation, but on the public JudgeBench benchmark even strong instruction-tuned judges ...
Biological neural circuits contain specialized substructures that support distinct computational functions, yet many bio-inspired neural networks borrow biologi...
This study significantly advances multi-texture synthesis using Neural Cellular Automata (NCAs) by introducing a novel training methodology that enables robust ...
Large Language Models (LLMs) have become increasingly prevalent in cloud-based platforms, propelled by the introduction of AI-based consumer and enterprise serv...
Agentic AI is changing the way users get work done. Following the success of OpenClawhttps://www.nvidia.com/en-us/ai/build-a-claw/, the community is embracing n...
A nonsmooth set-gradient ascent method is developed for moving finite approximation sets toward the Pareto front in multiobjective optimization. The method opti...
Asynchronous stochastic gradient descent (ASGD) is a standard way to exploit heterogeneous compute resources in distributed learning: instead of forcing fast wo...
Generative recommendation (GR) has emerged as a promising paradigm that replaces fragmented, scenario-specific architectures with unified Transformer-based mode...
Many real-world optimization problems are not naturally homogeneous vectors but composite design objects with heterogeneous parameters: integers, real values, B...
Computational social choice and algorithmic decision theory offer rich aggregation theory but no end-to-end, polynomial-time process for egalitarian self-govern...
Foundation models have transformed automated code generation, yet autonomous software-engineering agents remain unreliable in realistic development settings. Th...
Neural preconditioners for real-time physics simulation offer promising data-driven priors, but they often fail to capture long-range couplings efficiently beca...
When I joined the Codex engineering team in September 2025, Codex for Windows didn’t have a sandbox implementation meaning that Windows users were forced to cho...
Large-scale multi-objective optimization remains challenging because high-dimensional decision spaces, complex variable interactions, and limited function evalu...
Biological neural networks (BNNs) have been established as a powerful and adaptive substrate that offer the potential for incredibly energy and data efficient i...
As Large Language Models (LLMs) are transforming software development, the functional quality of generated code has become a central focus, leaving readability,...
Spiking Neural Networks (SNNs) are a promising framework for event-driven temporal processing. Prior work has improved temporal modeling through richer neuron d...
!https://cdn.platum.kr/wp-content/uploads/2026/05/dd-1024x576.jpg 프로그램 개요 스타트업 성장 파트너 디캠프가 일본 최대 철도·도시개발 기업 동일본여객철도JR동일본와 협력해 국내 스타트업에게 일본 스마트시티 프로젝트 실증 기회를 제공한...
Advances in large language models (LLMs) have recently opened new and promising avenues for small-molecule drug discovery. Yet existing LLM-based approaches for...
Computer-use agents (CUAs) automate on-screen work, as illustrated by GPT-5.4 and Claude. Yet their reliability on complex, low-frequency interactions is still ...
Recent large vision-language models (VLMs) remain fundamentally constrained by a persistent dichotomy: understanding and generation are treated as distinct prob...
Reconstructing the absolute 3D pose and shape of the hands from the user's viewpoint using a single head-mounted camera is crucial for practical egocentric inte...
Autoregressive video generation aims at real-time, open-ended synthesis. Yet, cinematic storytelling is not merely the endless extension of a single scene; it r...
Visual perception connects high-level semantic understanding to pixel-level perception, but most existing settings assume that the decisive evidence for identif...
In this paper, we propose AlphaGRPO, a novel framework that applies Group Relative Policy Optimization (GRPO) to AR-Diffusion Unified Multimodal Models (UMMs) t...
Surface reconstruction with differentiable rendering has achieved impressive performance in recent years, yet the pervasive photometric ambiguities have strictl...
Long-term memory is crucial for agents in specialized web environments, where success depends on recalling interface affordances, state dynamics, workflows, and...
We introduce Pion, a spectrum-preserving optimizer for large language model (LLM) training based on orthogonal equivalence transformation. Unlike additive optim...
Vision Transformers (ViTs) achieve strong data-driven scaling by leveraging all-to-all self-attention. However, this flexibility incurs a computational cost tha...
We explore the effectiveness of an LLM-guided query refinement paradigm for extending the usability of embedding models to challenging zero-shot search and clas...
Large language models (LLMs) are trained for downstream tasks by updating their parameters (e.g., via RL). However, updating parameters forces them to absorb ta...
In settings where labeled verifiable training data is the binding constraint, each checked example should be allocated carefully. The standard practice is to us...
Computer Use Agents (CUAs) can act through both atomic GUI actions, such as click and type, and high-level tool calls, such as API-based file operations, but th...
Recent advances in joint audio-video generation have been remarkable, yet real-world applications demand strong per-modality fidelity, cross-modal alignment, an...
LLM-based agents increasingly operate in persistent environments where they must store, update, and reason over information across many sessions. While prior be...
Sparse Mixture-of-Experts (SMoE) models enable scaling language models efficiently, but training them remains challenging, as routing can collapse onto few expe...
Reinforcement learning with verifiable rewards has enabled strong post-training gains in domains such as math and coding, though many open-ended settings rely o...
We introduce KV-Fold, a simple, training-free long-context inference protocol that treats the key-value (KV) cache as the accumulator in a left fold over sequen...
Looped Transformers offer a promising alternative to purely feed-forward computation by iteratively refining latent representations, improving language modeling...
Extreme weather and volatile wholesale electricity markets expose residential consumers to catastrophic financial risks, yet demand response at the distribution...
The continued improvements in language model capability have unlocked their widespread use as drivers of autonomous agents, for example in coding or computer us...
