PaperHub
影响力指数
91.47/100
前 0.5%
全站排名 #304
发表论文53 篇
平均评分5.0
年均产出17.7 篇/年

Tom Goldstein

Full Professor@University of Maryland, College Park·美国·OpenReview
研究方向

Gaussian Splatting · Vision-language models · Large language models · visualization · Computer vision · adversarial machine learning

6.5
15

Zebra-CoT: A Dataset for Interleaved Vision-Language Reasoning

ICLR 2026Poster
5.6
22

FictionalQA: A Dataset for Studying Memorization and Knowledge Acquisition

ICLR 2026Poster
5.5
21

DynaGuard: A Dynamic Guardian Model With User-Defined Policies

ICLR 2026Poster
通讯
5.0
23

Benchmarking Correctness and Security in Multi-Turn Code Generation

ICLR 2026Rejected
5.0
27

Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrence

ICLR 2026Rejected
4.8
19

Identifying and Evaluating Inactive Heads in Pretrained LLMs

ICLR 2026Poster
4.5
14

RL Is a Hammer and LLMs Are Nails: A Simple Reinforcement Learning Recipe for Strong Prompt Injection

ICLR 2026Rejected
4.5
5

CinePile: A Large-Scale Video Question Answering Dataset and Benchmark

ICLR 2026Withdrawn
通讯
4.0
5

BLIP3-o: A Family of Fully Open Unified Multimodal Models—Architecture, Training and Dataset

ICLR 2026Withdrawn
4.0
7

AegisLLM: Scaling Agentic Systems for Self-Reflective Defense in LLM Security

ICLR 2026Rejected
4.0
10

PHORECAST: Enabling AI Understanding of Public Health Outreach Across Populations

ICLR 2026Rejected
4.0
11

Scalable Pretraining of Retrieval Models

ICLR 2026Rejected
通讯
3.3
5

MORSE: A Suite of Programmatically Controllable Multimodal Reasoning Environments with Steerable Difficulty

ICLR 2026Desk Rejected
3.2
24

When Can You Get Away with Low Memory Adam?

ICLR 2026Rejected
通讯
2.5
5

Practical Memorization Tests for Detecting Copyrighted Data in Large Language Models

ICLR 2026Withdrawn
通讯
8.2
17

Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach

NeurIPS 2025Spotlight
通讯
7.3
24

LiveBench: A Challenging, Contamination-Limited LLM Benchmark

ICLR 2025Spotlight
7.3
17

Dense Backpropagation Improves Training for Sparse Mixture-of-Experts

NeurIPS 2025Poster
6.8
19

Quantifying Cross-Modality Memorization in Vision-Language Models

NeurIPS 2025Poster
三作
6.8
24

Refusal Tokens: A Simple Way to Calibrate Refusals in Large Language Models

COLM 2025Poster
通讯
6.4
22

Gemstones: A Model Suite for Multi-Faceted Scaling Laws

NeurIPS 2025Poster
通讯
6.0
22

LoRI: Reducing Cross-Task Interference in Multi-Task Low-Rank Adaptation

COLM 2025Poster
通讯
6.0
28

Has My System Prompt Been Used? Large Language Model Prompt Membership Inference

ICLR 2025Rejected
通讯
5.5
19

Dense Backpropagation Improves Routing for Sparsely-Gated Mixture-of-Experts

ICLR 2025Rejected
5.3
31

CinePile: A Long Video Question Answering Dataset and Benchmark

ICLR 2025Rejected
通讯
4.7
16

Running Huge Context Windows On Tiny GPUs

ICLR 2025Rejected
通讯
4.3
5

What do we learn from inverting CLIP models?

ICLR 2025Withdrawn
通讯
4.2
6

Gradient-based Optimization of Dataset Mixtures

ICLR 2025Withdrawn
通讯
4.2
6

OPTune: Efficient Online Preference Tuning

ICLR 2025Withdrawn
4.2
32

Refusal Tokens: A Simple Way to Calibrate Refusals in Large Language Models

ICLR 2025Rejected
通讯
4.2
6

From Pixels to Prose: A Large Dataset of Dense Image Captions

ICLR 2025Withdrawn
通讯
3.8
11

GenQA: An Instruction Dataset of LLM Generated Questions and Answers

ICLR 2025Withdrawn
通讯
3.7
4

Data Brittleness Estimation with Self-Supervised Features

ICLR 2025Withdrawn
通讯
3.7
10

A Simple Baseline for Predicting Future Events with Auto-Regressive Tabular Transformers

ICLR 2025Rejected
3.5
16

LLMs Boost the Performance of Decision Trees on Tabular Data across Sample Sizes

ICLR 2025Rejected
-1

Extracting Parameter Counts from Fine-Tuning APIs

ICLR 2025Withdrawn
通讯
合作者 (20)