開発者ツール B
418 件中 271〜300 件目を表示
-
Generation or Judgement? A Paradigm Perspective on LLM-Based Emotion-Cause Pair Extraction in Conversation
-
Feature Bagging Provides Stability
-
Breaking the Curse with BAND: Nonparametric Distribution Estimation in High Dimensions
-
Surrogate assisted diversity estimation in neural ensemble search
-
What Does It Take to Detect an AI Agent? Minimal Feature Sets for Behavioral Detection under Browser Automation
-
Siobahn Day Grady Wants Everyone to Be AI LiterateNCCU のグレイディ氏、HBCU 初の AI 研究所設立で AI リテラシー普及へIEEE Spectrumが、ノースカロライナ中央大学（NCCU）准教授シオバン・デイ・グレイディ氏を紹介。同氏は2025年1月、歴史的黒人大学（HBCU）で初となるAI研究所『IAIER』を設立し、AIリテラシーの普及と、大学間で偏在するAI教育の資金・産業ネットワーク格差の是正を目指す。※本文は途中で途切れており、IAIERの具体的な活動内容は末尾未確認。
-
Same Evidence, Different Target: Decoding How Diagnostic Evidence Bears on Causal Questions from Language-Model States
-
Latent-IM: Latent Interaction Management for Speech LLMs
-
Actions Have Consequences: Detecting Outcome Performativity using Intervention Testing
-
From Passive Video to Editable Experience: Physically Grounded Experience Synthesis for Embodied Intelligence
-
Human diversity fuels collective creativity that large language models cannot simulate or sustain
-
Bain Capital, Tillman Global invest $1.5bn in Eaton Fiber to support Verizon's fiber pushBain Capital と Tillman Global、Eaton Fiber に 15 億ドル出資DatacenterDynamics が、Bain Capital と Tillman Global が Eaton Fiber に 15 億ドルを出資したと報じた。Eaton Fiber は Verizon の光ファイバー網拡張を支える立場にあり、今回の資金がその推進を後押しするという。出資比率や資金使途の詳細は excerpt 外で未確認。
-
Think Short, Defer Smart, Act, and Repeat: Calibrated Reasoning and Uncertainty-Aware Deferral for Edge LLM Agents
-
Amortized Moment Matching for Visual Generation
-
No Data Is Not No Risk: Visibility Aware Graph-Based Inference of Business Conduct Risk
-
TREA-Net: A Transferable Residual Epidemiological Adaptation Network for Dengue Incidence Forecasting
-
From Representations to Behaviors: Exploring the Person-Situation-Behavior Triad in LLMs
-
AI as Friction for Reflection Support in Ideation
-
FedTopo: Relation-Level Topology Sharing for Model-Heterogeneous Federated Learning
-
When Does Span-Guided Detoxification Help? Human Preferences and Evaluator Diagnostics in a Controlled Comparison
-
SecRespond: Benchmarking AI Agents for Real-World Post-Compromise Incident Response
-
Property-driven Causal Abstractions for Markov Decision Processes
-
Do Latent Channels Actually Communicate? A Causal Audit of Latent Multi-Agent LLM
-
MediaWiki Code2Code Search: Neural Retrieval for the Semantic Discovery of Open-Source Software Entities
-
Relation Geometry in Semantic Space of Language Models
-
AI Hyper-Scaling Digital InequalityAI の急拡大が既存のデジタル格差を増幅、南ア・インドネシアが代替を模索IEEE Spectrum の寄稿。AI がメール・コード作成、採用選考、教育・医療・行政などの日常インフラに組み込まれ「万人のための AI」が語られる一方、筆者は欧州・サブサハラアフリカ・東南アジアでのデジタル包摂の経験から、新技術が接続性・スキル・制度能力で既に階層化した社会に着地し格差をむしろ増幅すると論じる。米中が支配するフロンティアモデル競争を直接なぞらない参加の道として南アフリカとインドネシアの事例に触れる（本文は途中まで、具体的施策は抜粋外）。
-
An Attention-Based Framework for Alzheimers Disease Classification Using Resting-State fMRI
-
Multimodal fusion of visual and morphometric features for avian bone classification
-
Zero-Shot Face-to-Speech Synthesis via Latent Space Adaptation of a Style-Diffusion TTS Model
-
Dual Inversion for Text-to-Image Diffusion Models: From Both Prompt and Noise Perspectives