paper.dinq.meAI 与前沿科学顶会 · 学术新星雷达
登录 DINQ
返回新星榜
JL

Jaeho Lee

ACL 2026×2ICML 2026数据与预训练理论与学习动力学对齐与偏好效率与系统多模态
新星分资深
17
一作贡献0
核心/参与5
方向广度13
合作网络0
0 一作 · 3 篇 · ACL 2026 / ICML 2026

获取完整研究档案

首次获取该候选人的完整档案及 StartLux 评估共 20 积分;每位用户单独解锁,之后查看免费。主动刷新 20 积分。

联系方式

实时查找电子邮箱,费用不超过20积分;未找到邮箱不扣费。

2026 顶会论文

3
作者位次 #3/3ICML 2026Poster· Deep Learning / Theory

Over-Alignment vs Over-Fitting: The Role of Feature Learning Strength in Generalization

Taesun Yeom · Taehyeok Ha · Jaeho Lee

理论与学习动力学对齐与偏好OpenReview
作者位次 #3/3ACL 2026Poster· Natural Language Processing

Speculative End-Turn Detector for Efficient Speech Chatbot Assistant

Hyunjong Ok · Suho Yoo · Jaeho Lee

摘要

Spoken dialogue systems powered by large language models have demonstrated remarkable abilities in understanding human speech and generating appropriate spoken responses.However, these systems struggle with end-turn detection (ETD)—the ability to distinguish between user turn completion and hesit...

效率与系统多模态数据与预训练OpenReview 项目页
作者位次 #2/2ACL 2026Findings· Findings of ACL

Lost in the Prompt Order: Revealing the Limitations of Causal Attention in Language Models

Hyunjong Ok · Jaeho Lee

摘要

Large language models exhibit surprising sensitivity to the structure of the prompt, but the mechanisms underlying this sensitivity remain poorly understood. In this work, we conduct an in-depth investigation on a striking case: in multiple-choice question answering, placing context before the qu...

可解释性数据与预训练检索与知识OpenReview 项目页

常见合作者