MLINFO COLLECTION
日次更新の新着・公開記事を新しい順に掲載しています。
Self-distillation methods for large language models with an exponential moving average (EMA) teacher already retain a history of student updates. Muon nevertheless stores a separate momentum buffer,…
An evaluated candidate can improve a local search step, alter the population used by later steps, and affect a solver's final result. We distinguish these three questions using Da Gou Gun…
On-policy reinforcement learning for large language models assumes that training responses are sampled from the current policy. In practice, an inference engine samples the responses and a separate…
Training acceleration methods for diffusion transformers, such as REPA and dispersive regularization, are commonly explained by additional supervision, spatial-structure priors, or improved feature…
As publications and products on time series forecasting (TSF) proliferate, researchers face a growing burden of analyzing existing work. First, not all publications release code.
Across coding, web, and tool-use tasks, language-model agents are now routinely trained with the same recipe: supervised fine-tuning (SFT) to learn the interface, then multi-turn reinforcement…
Achieving the broad task generalization of embodied models in physical environments remains a central challenge in robot learning. We study this challenge in World Action Models (WAMs) through two…
A capable voice assistant must do more than recognize words; it must understand requests, reason over context, sustain conversations, and act through tools. We present Apriel Voice 27B Preview, an…
A voice agent's speech recognizer transcribes everyone near the microphone, including people other than the caller. I fine-tuned a public causal target-speaker extraction model (REAL-TSE, 27M…
An observer sees only part of the surrounding world, yet must keep track of what lies beyond the current view. Spatial perception therefore involves more than recognizing individual images: it calls…
Interactive world models generate visual content conditioned on user actions, yet maintaining spatial consistency over long-horizon exploration in real time remains challenging. Existing approaches…
The hair of stylized characters in games and avatars is built from hair cards: artist-placed ribbons whose geometry is the visible hair, animated by spring bones. Existing image-based methods produce…
Foundation models are typically trained in stages: pretraining, midtraining, supervised fine-tuning (SFT), and reinforcement learning (RL), with changes in both data and learning algorithms. We show…
Do research agents have taste, the judgment of which experiments are worth running before their outcomes are known? We give six frontier models seven recent NLP papers.
Retrieval-augmented generation commonly serves separate models for reranking and answering, duplicating model residency and passage processing. We introduce Glance, which equips the generator with a…
We present PhysInsert, a framework for physically plausible reference-based object insertion that jointly adapts objects and their surrounding scenes. Existing methods often rely on masked inputs,…
Vision-Language-Action (VLA) models have demonstrated promising capabilities in robotic manipulation by integrating visual perception, language understanding, and action generation. Recent advances…
Weight-normalized (WN) training is emerging as a compelling alternative to conventional weight decay (WD), with recent studies reporting substantial pretraining loss gains. We show that these gains…
Recent advances in video world models have enabled the generation of increasingly realistic and interactive virtual environments, opening new possibilities for player-customizable gameplay. Agentic…
Large language model (LLM)-based multi-agent systems (MAS) coordinate specialized agents to solve complex tasks, turning each user request into interdependent model and tool operations. Unlike…
To succeed in real-world long-horizon software development, coding agents must satisfy multi-dimensional task specifications, such as interface compatibility and error handling, and do so without…
Video generation is shifting from end-to-end synthesis toward agentic generation, where planning, tool use, and feedback enable coherent, controllable multi-shot videos. In parallel, video world…
Agentic image editors decompose complex requests into steps, but when a step fails, they mostly retry with the same narrow set of tools and a judge that scores each output by eye directly. We argue…
Video diffusion transformers generate high-quality videos, but the quadratic cost of 3D full attention dominates their inference. Dynamic block-sparse attention reduces this cost by computing exact…