AI Wiki — Knowledge Base

5月25日、内閣総理大臣官邸で開催された車座対談に、Sakana AIの伊藤錬が参加しました。金融など各産業に特化した高度なAI実装の実例をご紹介しました。また、海外の先端モデルも活用しつつ、日本独自の技術で、我が国の防衛の自律性やデータ主権を確保する具体的方法についても高市総…

5月25日、内閣総理大臣官邸で開催された車座対談に、Sakana AIの伊藤錬が参加しました。金融など各産業に特化した高度なAI実装の実例をご紹介しました。また、海外の先端モデルも活用しつつ、日本独自の技術で、我が国の防衛の自律性やデータ主権を確保する具体的方法についても高市総理と貴重な意見交換の機会をいただきました。本日、スタートアップ４社（Atomis、Sakana AI、Oceanic

researchdavid-ha--x

25 May 2026

research

The Quantization Benefits of Residual-Free Transformers

The Telegraph: Macron tore up 65 years of doctrine to defend Europe with French nukes, with or without the US. Poland, Germany, the Netherla…

The Timing Dependencies of Trust: Speed, Accuracy, and cBCI Neuro-Decoupling in Human-AI Teams

The Tokenizer Tax Across 25 European Languages: Domain Invariance, Cross-Lingual Few-Shot Effects, and the Ukrainian Penalty

Theoretical Analysis of Sparse Optimization with Reparameterization, Weight Decay, and Adaptive Learning Rate

They Are Not the Same: Direct Causes Are Not Grounded Emotion Explanations

TIGER: Text-Informed Generalized Enzyme-Reaction Retrieval

.@tinkerapi by TML is a highly underrated product. It should be part of the “starter post training” infra for any newco

TinyFormer: Preserving Tiny Objects in YOLO-DETRHybridReal-time Detectors

Towards end-to-end LLM-based censoring-aware survival analysis

Towards Evaluation Engineering: An Empirical Study of ML Evaluation Harnesses in the Wild

Towards Long-Horizon Interpretability: Efficient and Faithful Multi-Token Attribution for Reasoning LLMs

Toxicity in Twitch Chats: An LLM-Based Analysis Across Gaming Communities

TRAFA: Anticipating User Actions to Reduce Errors in Procedural Tasks with Predictive Feedback

Train-Free Segmentation in MRI with Cubical Persistent Homology

Trained quantum neural networks are Gaussian processes

Trajectory-Based Difficulty Scoring for Reliable Learning on Tabular Data

Transformer-based few-shot learning for modeling Electricity Consumption Profiles with minimal data across thousands of domains

Triplet-Block Diffusion RWKV

TSFLora: Token-Compressed Split Fine-Tuning for Wireless Edge Networks

TUBE: Tangent Upper Bound on Evidence for Discrete Diffusion Language Models

Turn-Based Structural Triggers: Prompt-Free Backdoors in Multi-Turn LLMs

Uncovering Autoregressive LLM Knowledge of Thematic Fit in Event Representation

Uncovering Vulnerabilities of LLM-Assisted Cyber Threat Intelligence

Understanding, Accelerating, and Improving MeanFlow Training

Understanding the Impact of Geometric Foundation Models on Vision-Language-Action Models

Universal Activation Verbalizer: A Unified Framework for Cross-Model Activation Explanation

Variable Clustering via Distributionally Robust Nodewise Regression

Verified SHAP: Provable Bounds for Exact Shapley Values of Neural Networks

Volatility Surface Reconstruction using Deep Learning under No-Arbitrage Constraints

Voting with the Graph: Stable RLAIF via Topological Consistency Maximization

What Questions Should Robots Be Able to Answer? A Dataset of User Questions for Explainable Robotics

When Does Synthetic Patent Data Help? Volume-Fidelity Trade-offs in Low-Resource Multi-Label Classification

When Gradients Collide: Failure Modes of Multi-Objective Prompt Optimization for LLM Judges

When In-Distribution Gains Fail: Evaluating Weak-to-Strong Reward Models under Preference Shift

When Interpretability Becomes a Liability: Adversarial Attacks on CBM Concept Layers

Which Is Better For Reducing Outdated and Vulnerable Dependencies: Pinning or Floating?

WINO: A Weak-Form Physics Informed Neural Operator for Hyperelasticity on Variable Domains

Word Class Representations Spontaneously Emerge from Successor Representations Trained on Natural Language

WTKO-CNN: Deep Learning Reveals Sequence Motifs Distinguishing Wild-Type and Knockout ATAC-seq Peaks

You Can Ground Earlier than See: An Effective and Efficient Pipeline for Temporal Sentence Grounding in Compressed Videos

Zero-Shot Parkinson's Disease Detection from Speech: Comparing Large Audio and Language Models

Zeroth-Order Nonconvex Nonsmooth Optimization with Heavy-Tailed Noise

A comprehensive evaluation of pretraining strategies for channel-agnostic contrastive self-supervision of biosignals

A drone-based framework for coral habitat mapping via weakly supervised segmentation

A Fine-Tuned BERT Classifier for Personal-Letter Titles in Late-Ming and Early-Qing Collected Works

A graph-based analysis of semantic types and coercion in contextualized word embeddings