今週のストーリー

NVIDIA × 学習・ファインチューニング

NVIDIA、Physical AI向けBEV Pooling高速化

z = 5.1 7社横断 94本 2026-06-25（±1日窓） /ja/ai/story/nvidia-2026-06-25

✎ ストーリー本文

NVIDIA が Physical AI 向けに BEV（Bird's Eye View）Pooling を GPU 上で高速化する実装を公開。自律走行やロボット制御で頻用される表現の裏側を薄く速くする話で、学習・fine-tune の効率化が主題。Hugging Face 側からは Transformers Fine-Tuning を NeMo AutoModel で速める話や、Transformers.js の Cross-Origin Storage 実験が同居し、「学習パイプラインを短く軽くする」動きが横断していた。派手なモデル発表がない中で、fine-tuning の実務が着実に整備される様子が読み取れる。

▲ 公式・報道

公式

Accelerating BEV Pooling on NVIDIA GPUs for Physical AI Applications

NVIDIA Developer Blog ・ 2026-06-24 ・ 📌

NVIDIA、自動運転・ロボット向けに GPU 上の BEV プーリングを高速化

報道

中国が人型ロボット開発で急成長しているワケ　日本が学ぶべきポイントは？　専門家が解説

ITmedia AI+ ・ 2026-06-25

中国が人型ロボット開発で急成長する理由と日本が学ぶべき点を専門家が解説

報道

「今日言うつもりはなかったが……」　孫正義氏が明かした「ロボット自動量産工場」の実態

ITmedia AI+ ・ 2026-06-25

孫正義氏、株主総会で「ロボット自動量産工場」の実態に言及

公式

Accelerating Transformers Fine-Tuning with NVIDIA NeMo AutoModel

Hugging Face Blog ・ 2026-06-24

Hugging Face、NVIDIA NeMo AutoModelで微調整を高速化

報道

「Transformerの最大475倍」　富士通、GPUを効率的に使うLLMアーキテクチャ「PHOTON」開発

ITmedia AI+ ・ 2026-06-24

富士通、GPUを効率利用するLLMアーキテクチャ「PHOTON」を開発

報道

陸自駐屯地で四足歩行型の警備用ロボットが見回り　GMOインターネットグループが開発

ITmedia AI+ ・ 2026-06-24

GMOインターネットG、四足歩行型の警備ロボットを開発し陸自で導入検証

報道

“中国ヒューマノイド革命”はなぜ起きた、異業種や大手テックが動かす市場の今

ITmedia AI+ ・ 2026-06-23

中国ヒューマノイドロボット、出荷7倍・世界シェア8割で量産期へ

報道

トヨタ系金融会社はなぜ「AIエージェントだけ」でも「RPAだけ」でもなく“併用”にしたのか

ITmedia AI+ ・ 2026-06-23

トヨタファイナンス、問い合わせ対応にAIエージェントとRPAの併用を導入

公式

Experimenting with the proposed Cross-Origin Storage API in Transformers.js

Hugging Face Blog ・ 2026-06-23

Hugging Face、Transformers.jsで提案中のCross-Origin Storage APIを検証

コミュニティ

Porting the Moebius 0.2B image inpainting model to run in the browser with Claude Code

Simon Willison's Weblog ・ 2026-06-22

Simon Willison、0.2B 画像補完モデル Moebius を Claude Code でブラウザ移植

学術（arxiv ほか） 84本 ▾

学術

Language-Based Digital Twins for Elderly Cognitive Assistance

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-25

学術

Hallucination in World Models is Predictable and Preventable

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

A Multi-Fidelity Convolutional Autoencoder-Transfer Learning Framework for Guided-Wave-Based Damage Diagnosis Using Large Simulated and Limited Experimental Datasets

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

E-TTS: A New Embodied Test-Time Scaling Framework for Robotic Manipulation

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-25

学術

Paved with True Intents: Intent-Aware Training Improves LLM Safety Classification Across Training Regimes

arXiv cs.CL (Computation and Language) ・ 2026-06-25

学術

Graph Neural Networks Applications Across Domains: All Insights You Need

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

HarmVideoBench: Benchmarking Harmful Video Understanding in Large Multimodal Models

arXiv cs.CL (Computation and Language) ・ 2026-06-25

学術

Learning to Fold: prizewinning solution at LeHome Challenge 2026 (1st place online, 2nd offline)

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

Safe Autoregressive Image Generation with Iterative Self-Improving Codebooks

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-25

学術

Kolmogorov Arnold networks (KAN) for aerodynamic prediction: a comparison with MLPs and GNNs

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

Transformer-Based Classification of Bacterial Raman Spectra with LOOCV

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

Inherited Circuits, Learned Semantics: How Fine-Tuning Creates Evasion Vulnerabilities Invisible to Standard Evaluation

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-25

学術

Towards Explainable Adjudicative Variance: Quantifying Judicial Discretion via Gated Multi-Task Learning

arXiv cs.CL (Computation and Language) ・ 2026-06-25

学術

Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization

arXiv cs.CL (Computation and Language) ・ 2026-06-25

学術

Just how sure are you? Improving Verbalized Uncertainty Calibration in Medical VQA

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

GAVEL: Grounded Caption Error Verification and Localization

arXiv cs.CL (Computation and Language) ・ 2026-06-25

学術

SamaVaani: Auditing and Debiasing Multilingual Clinical ASR for Indian Languages

arXiv cs.CL (Computation and Language) ・ 2026-06-25

学術

Reasoning Quality Emerges Early: Data Curation for Reasoning Models

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

AIGP: An LLM-Based Framework for Long-Term Value Alignment in E-Commerce Pricing

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

Escaping Iterative Parameter-Space Noise: Differentially Private Learning with a Hypernetwork

arXiv cs.LG (Machine Learning) ・ 2026-06-25

学術

Learning Action Priors for Cross-embodiment Robot Manipulation

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

How Robust is OCR-Reasoning? Evaluating OCR-Reasoning Robustness of Vision-Language Models under Visual Perturbations

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Detect, Unlearn, Restore: Defending Text Summarization Models Against Data Poisoning

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Privacy Vulnerabilities of Attention Layers in Tabular Foundation Models and Protection of High-Risk Queries

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

The Tatoxa System for Text Detoxification in Low-Resource Languages: The Case of Tatar

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

FORCE: Efficient VLA Reinforcement Fine-Tuning via Value-Calibrated Warm-up and Self-Distillation

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

Dziri Voicebot: An End-to-End Low-Resource Speech-to-Speech Conversational System for Algerian Dialect

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Hierarchical Reinforcement Learning for Neural Network Compression (HiReLC): Pruning and Quantization

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

Weave of Formal Thought

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

Multi-Agent Goal Recognition with Team- and Goal-Conditioned Reinforcement Learning and Factorized Branch-and-Bound

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

Tensorion: A Tensor-Aware Generalization of the Muon Optimizer

arXiv cs.LG (Machine Learning) ・ 2026-06-24

学術

WinDOM: Self-Family Distillation for Small-Model GUI Grounding

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

Enhancing Brain MRI Anomaly Detection and Reasoning with ROI Rethink and Synthetic Data

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

AutoRelAnnotator: Calibrated Model Cascades for Cost-Efficient Relevance Evaluation in Sponsored Search

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

Edges Before Embeddings: A Confidence-Aware Blur Gate for Vision-Language Pipelines

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

ROAD-VLA: Robust Online Adaptation via Self-Distillation for Vision-Language-Action Models

arXiv cs.LG (Machine Learning) ・ 2026-06-24

学術

Uncertainty Quantification for Computer-Use Agents: A Benchmark across Vision-Language Models and GUI Grounding Datasets

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

Memory-Efficient Policy Libraries with Low-Rank Adaptation in Reinforcement Learning

arXiv cs.LG (Machine Learning) ・ 2026-06-24

学術

BitNet Text Embeddings

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Steering Vision-Language Models with Joint Sparse Autoencoders

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-24

学術

TL++: Accuracy and Privacy Preserving Traversal Learning for Distributed Intelligent Systems

arXiv cs.LG (Machine Learning) ・ 2026-06-24

学術

Expresso-AI: Explainable Video-Based Deep Learning Models for Depression Diagnosis

arXiv cs.LG (Machine Learning) ・ 2026-06-24

学術

Riazi-8B: An Urdu Large Language Model for Mathematical Reasoning

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Fault of Our Stars: Behavioral Drivers of Rating-Sentiment Incongruence

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Spam and Sentiment Detection in Arabic Tweets Using MARBERT Model

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Optimizing Abstractive Summarization With Fine-Tuned PEGASUS

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

The Generalization Spectrum: A Chromatographic Approach to Evaluating Learning Algorithms

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

Evaluating Japanese Dialect Robustness Across Speech and Text-based Large Language Models

arXiv cs.CL (Computation and Language) ・ 2026-06-24

学術

InSight: Self-Guided Skill Acquisition via Steerable VLAs

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

FLUX3D: High-Fidelity 3D Gaussian Generation with Diffusion-Aligned Sparse Representation

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

Matching Tasks to Objectives: Fine-Tuning and Prompt-Tuning Strategies for Encoder-Decoder Pre-trained Language Models

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

L3Cube-MahaPOS: A Marathi Part-of-Speech Tagging Dataset and BERT Models

arXiv cs.CL (Computation and Language) ・ 2026-06-23

学術

SHERLOC: Structured Diagnostic Localization for Code Repair Agents

arXiv cs.CL (Computation and Language) ・ 2026-06-23

学術

OrbitForge: Text-to-3D Scene Generation via Reconstruction-Anchored Video Synthesis

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

UniDrive: A Unified Vision-Language and Grounding Framework for Interpretable Risk Understanding in Autonomous Driving

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

Can Scale Save Us From Plasticity Loss in Large Language Models?

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

Decentralised AI Training and Inference with BlockTrain

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

A Physics-Informed Fourier-Wavelet Transformer for Multiscale Computational Fluid Dynamics Surrogate Modeling

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

CineCap: Structured Reasoning with Spatio-Temporal Anchors for Cinematographic Video Captioning

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

ScaleToT: Generalizing Structured LLM Reasoning for Billion-Scale Low-Activity User Modeling

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

Uncertainty-Aware Longitudinal Forecasting of Alzheimer's Disease Progression Using Deep Learning

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

Qwen-AgentWorld: Language World Models for General Agents

arXiv cs.CL (Computation and Language) ・ 2026-06-23

学術

Reinforcement Learning for Computer-Use Agents with Autonomous Evaluation

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-23

学術

An LLM-based Two-Stage Transformer Framework for Cross-Domain Bearing Fault Diagnosis with Limited Data

arXiv cs.CL (Computation and Language) ・ 2026-06-23

学術

MedPCFM: Improving Medical Point Cloud Completion by Integrating Point Transformers and Flow Matching

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

Parallel Manifold Steering: Efficient Adaptation of Large Associative Memories via Residual Energy Shaping

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

PHANTOM: A Large-Scale Dataset of Multimodal Adversarial Attacks for Vision-Language Models

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

AutoSpecNER: A Fine-Grained Named Entity Recognition Dataset for Vehicle Specification Extraction

arXiv cs.CL (Computation and Language) ・ 2026-06-23

学術

Open-Vocabulary BEV Segmentation with 3D-Aware Geometric Constraints

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

SURGELLM: Rethinking Multi-Task Evaluation through Task-Aware Feature Gating with Class-Balanced Normalization

arXiv cs.CL (Computation and Language) ・ 2026-06-23

学術

Automated Residual Plot Assessment With the R Package autovi and the Shiny Application autovi.web

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

Lightweight Transformer Models for On-Device Fault Detection: A Benchmark Study on Resource-Constrained Deployment

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

When Top-1 Fails: Calibrating LoRA Monitors for Masked Diffusion LMs

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

NeuroSonic: Conditional Flow Matching for EEG-to-Speech Reconstruction

arXiv cs.LG (Machine Learning) ・ 2026-06-23

学術

Tapered Language Models

arXiv cs.LG (Machine Learning) ・ 2026-06-22

学術

Muown Implicitly Performs Angular Step-size Decay

arXiv cs.LG (Machine Learning) ・ 2026-06-22

学術

DiT-Reward: Generative Representations for Text-to-Image Reward Modeling

arXiv cs.LG (Machine Learning) ・ 2026-06-22

学術

Scaling Linear Mode Connectivity and Merging to Billion Parameter Pretrained Transformers

arXiv cs.LG (Machine Learning) ・ 2026-06-22

学術

The Topology of Ill-Posed Questions: Persistent Homology for Detection and Steering in LLMs

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-22

学術

A Generative Model for Closed-Loop Microsimulation of Signalized Intersections

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-22

学術

The Energy Consumption of Transformer Fine-Tuning: A Roofline-Inspired Scaling Model

arXiv cs.LG (Machine Learning) ・ 2026-06-22

学術

HyperQuant: A Rate-Distortion-Optimal Quantization Pipeline for Large Language and Diffusion Models

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-22

学術

Energy-Based Transformers as Predictors of Reading Difficulty

arXiv cs.AI (Artificial Intelligence) ・ 2026-06-22

← ストーリーアーカイブ