activity
20212026
most citedAHP-Powered LLM Reasoning for Multi-Criteria Evaluation of Open-Ended Responses

1 citations · 1 across the 17 of their papers we have counts for

collaborators
Showing cs.LGShow all

7 papers · 1 filter

cs.LG2026

Disentangling Steering Vectors

Takeru Hiramatsu, Kyohei Atarashi, Koh Takeuchi +1

Activation steering has emerged as a lightweight, inference-time approach to control the behavior of Large Language Models (LLMs). However, traditional steering vectors used to int…

cs.LG2026

Long-Term Outlier Prediction Through Outlier Score Modeling

Yuma Aoki, Joon Park, Koh Takeuchi +6

This study addresses an important gap in time series outlier detection by proposing a novel problem setting: long-term outlier prediction. Conventional methods primarily focus on i…

cs.LG2025

Robust Anomaly Detection Under Normality Distribution Shift in Dynamic Graphs

Xiaoyang Xu, Xiaofeng Lin, Koh Takeuchi +2

Anomaly detection in dynamic graphs is a critical task with broad real-world applications, including social networks, e-commerce, and cybersecurity. Most existing methods assume th…

cs.LG2025

Unpacking the Implicit Norm Dynamics of Sharpness-Aware Minimization in Tensorized Models

Tianxiao Cao, Kyohei Atarashi, Hisashi Kashima

Sharpness-Aware Minimization (SAM) has been proven to be an effective optimization technique for improving generalization in overparameterized models. While prior works have explor…

cs.LG2025

Dynamic Feature Selection from Variable Feature Sets Using Features of Features

Katsumi Takahashi, Koh Takeuchi, Hisashi Kashima

Machine learning models usually assume that a set of feature values used to obtain an output is fixed in advance. However, in many real-world problems, a cost is associated with me…

cs.LG2024

Treatment Effect Estimation for Graph-Structured Targets

Shonosuke Harada, Ryosuke Yoneda, Hisashi Kashima

Treatment effect estimation, which helps understand the causality between treatment and outcome variable, is a central task in decision-making across various domains. While most st…