AI Wiki — Knowledge Base

音声AIの素早さと賢さを両立できるか？私たち人間は会話の中で、言いたいことを全部まとめてから話し始めるのではなく、話しながら考えを整理していきます。応答の速い Speech-to-Speech モデルは、この「話しながら考える」を実現しましたが、そのぶん思考が浅くなりがちです。…

音声AIの素早さと賢さを両立できるか？私たち人間は会話の中で、言いたいことを全部まとめてから話し始めるのではなく、話しながら考えを整理していきます。応答の速い Speech-to-Speech モデルは、この「話しながら考える」を実現しましたが、そのぶん思考が浅くなりがちです。かといって知識豊富な LLM を挟むカスケード型では、遅延が生じるため「話しながら」が成立しません。そこで Sakan

model-releasesdavid-ha--x

30 Apr 2026

Search: “papers”

Making Logic a First-Class Citizen in Generative ML for Networking

NashPG: A Policy Gradient Method with Iteratively Refined Regularization for Finding Nash Equilibria

Normativity and Productivism: Ableist Intelligence? A Degrowth Analysis of AI Sign Language Translation Tools for Deaf People

On the Proper Treatment of Units in Surprisal Theory

OpAgent: Operator Agent for Web Navigation

Optimization before Evaluation: Evaluation with Unoptimised Prompts Can be Misleading

People-Centred Medical Image Analysis

Policy-Governed LLM Routing with Intent Matching for Instrument Laboratories

Policy-Grounded Safety Evaluation of 20 Large Language Models

Pragmos: A Process Agentic Modeling System

PVeRA: Probabilistic Vector-Based Random Matrix Adaptation

Qualitative Evaluation of Language Model Rescoring in Automatic Speech Recognition

Quantum Masked Autoencoders for Vision Learning

Real-Time GPU-Accelerated Monte Carlo Evaluation of Safety-Critical AEB Systems Under Uncertainty

Rethinking Agentic Reinforcement Learning In Large Language Models

Revealing the Impact of Visual Text Style on Attribute-based Descriptions Produced by Large Visual Language Models

Robot Learning from Human Videos: A Survey

Robot Planning and Situation Handling with Active Perception

Robust Lightweight Crack Classification for Real-Time UAV Bridge Inspection

RopeDreamer: A Kinematic Recurrent State Space Model for Dynamics of Flexible Deformable Linear Objects

Self-Evolving Software Agents

State-Dependent Lyapunov Method for Rank-1 Matrix Factorization

Student Classroom Behavior Recognition Based on Improved YOLOv8s

Taming Noise-Induced Prototype Degradation for Privacy-Preserving Personalized Federated Fine-Tuning

Taxon: Hierarchical Tax Code Prediction with Semantically Aligned LLM Expert Guidance

Test Before You Deploy: Governing Updates in the LLM Supply Chain

The Effects of Visual Priming on Cooperative Behavior in Vision-Language Models

To Build or Not to Build? Factors that Lead to Non-Development or Abandonment of AI Systems

To Diff or Not to Diff? Structure-Aware and Adaptive Output Formats for Efficient LLM-based Code Editing

Towards Neuro-symbolic Causal Rule Synthesis, Verification, and Evaluation Grounded in Legal and Safety Principles

TripVVT: A Large-Scale Triplet Dataset and a Coarse-Mask Baseline for In-the-Wild Video Virtual Try-On

TypeBandit: Type-Level Context Allocation and Reweighting for Effective Attribute Completion in Heterogeneous Graph Neural Networks

Uncertainty Quantification Framework for Aerial and UAV Photogrammetry through Error Propagation

What Makes a Good Terminal-Agent Benchmark Task: A Guideline for Adversarial, Difficult, and Legible Evaluation Design

When Roles Fail: Epistemic Constraints on Advocate Role Fidelity in LLM-Based Political Statement Analysis

YOSE: You Only Select Essential Tokens for Efficient DiT-based Video Object Removal

A Multi-Dataset Benchmark of Multiple Instance Learning for 3D Neuroimage Classification

A Practice of Post-Training on Llama-3 70B with Optimal Selection of Additional Language Mixture Ratio

A Survey of Multi-Agent Deep Reinforcement Learning with Graph Neural Network-Based Communication

A Survey on the Safety and Security Threats of Computer-Using Agents: JARVIS or Ultron?

Action Hints: Semantic Typicality and Context Uniqueness for Generalizable Skeleton-based Video Anomaly Detection

Adaptive and Fine-grained Module-wise Expert Pruning for Efficient LoRA-MoE Fine-Tuning

Addressing Performance Saturation for LLM RL via Precise Entropy Curve Control

Affective Flow Language Model for Emotional Support Conversation

Agentic Search in the Wild: Intents and Trajectory Dynamics from 14M+ Real Search Requests

AHASD: Asynchronous Heterogeneous Architecture for LLM Adaptive Drafting Speculative Decoding on Mobile Devices

Asynchronous Federated Unlearning with Invariance Calibration for Medical Imaging