影响力指数

76.26/100

前 1.7%

全站排名 #1,124

发表论文14 篇

平均评分5.5

年均产出4.7 篇/年

Robin Jia

Assistant Professor@University of Southern California·美国·OpenReview

研究方向

question answering · robustness · large language models · uncertainty estimation · distribution shift · interpretability

7.5

Hubble: a Model Suite to Advance the Study of LLM Memorization

ICLR 2026Oral

通讯

5.5

Function Induction and Task Generalization: An Interpretability Study with Off-by-One Addition

ICLR 2026Poster

二作

5.5

Cancer-Myth: Evaluating Large Language Models on Patient Questions with False Presuppositions

ICLR 2026Poster

通讯

5.0

FoNE: Precise Single-Token Number Embeddings via Fourier Features

ICLR 2026Poster

4.5

When Do LLMs Admit Their Mistakes? Understanding the Role of Model Belief in Retraction

ICLR 2026Withdrawn

二作

3.5

When Do Transformers Learn Heuristics for Graph Connectivity?

ICLR 2026Desk Rejected

三作

3.3

Textual Steering Vectors Can Improve Visual Understanding in Multimodal Large Language Models

合作者 (20)

Robin Jia

Hubble: a Model Suite to Advance the Study of LLM Memorization

Function Induction and Task Generalization: An Interpretability Study with Off-by-One Addition

Cancer-Myth: Evaluating Large Language Models on Patient Questions with False Presuppositions

FoNE: Precise Single-Token Number Embeddings via Fourier Features

When Do LLMs Admit Their Mistakes? Understanding the Role of Model Belief in Retraction

When Do Transformers Learn Heuristics for Graph Connectivity?

Textual Steering Vectors Can Improve Visual Understanding in Multimodal Large Language Models

LLM Unlearning Without an Expert Curated Dataset

TLDR: Token-Level Detective Reward Model for Large Vision Language Models

Teaching Models to Understand (but not Generate) High-risk Data