Tom Goldstein's Lab at University of Maryland, College Park

university

http://www.cs.umd.edu/~tomg/

AI & ML interests

AI security & privacy, algorithmic bias, foundations of ML

Recent Activity

kaiyuyue updated a collection 10 days ago

kaiyuyue submitted a paper 10 days ago

Image Generation with a Sphere Encoder

kaiyuyue updated a collection 11 days ago

View all activity

Papers

Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrence

Gemstones: A Model Suite for Multi-Faceted Scaling Laws

View all Papers

updated a collection 10 days ago

Sphere Encoder

Image Generation with a Sphere Encoder • 3 items • Updated 5 days ago

submitted a paper to Daily Papers 10 days ago

Image Generation with a Sphere Encoder

Paper • 2602.15030 • Published 19 days ago • 15

updated a collection 11 days ago

Sphere Encoder

Image Generation with a Sphere Encoder • 3 items • Updated 5 days ago

updated a collection 15 days ago

MTP-LM

Models to accompany research paper on training multi token prediction language models using self-distillation. • 4 items • Updated 6 days ago • 3

submitted a paper to Daily Papers 16 days ago

NESSiE: The Necessary Safety Benchmark -- Identifying Errors that should not Exist

Paper • 2602.16756 • Published 18 days ago • 3

in tomg-group-umd/huginn-dataset 19 days ago

📋 Documentation Enhancement Suggestion

#4 opened 21 days ago by

updated a collection 22 days ago

MTP-LM

Models to accompany research paper on training multi token prediction language models using self-distillation. • 4 items • Updated 6 days ago • 3

updated a collection 25 days ago

MTP-LM

Models to accompany research paper on training multi token prediction language models using self-distillation. • 4 items • Updated 6 days ago • 3

updated a collection 26 days ago

MTP-LM

Models to accompany research paper on training multi token prediction language models using self-distillation. • 4 items • Updated 6 days ago • 3

updated a collection 26 days ago

MTP-LM

Models to accompany research paper on training multi token prediction language models using self-distillation. • 4 items • Updated 6 days ago • 3

in tomg-group-umd/Gemstone-512x13 2 months ago

Update README.md

#1 opened 2 months ago by

updated a collection 3 months ago

PixelProse

4 items • Updated 10 days ago

updated a dataset 3 months ago

tomg-group-umd/pixelprose

Viewer • Updated Dec 13, 2025 • 15.6M • 467 • 163

authored a paper 4 months ago

Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrence

Paper • 2511.07384 • Published Nov 10, 2025 • 19

authored a paper 4 months ago

Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrence

Paper • 2511.07384 • Published Nov 10, 2025 • 19

updated a collection 4 months ago

Retrofitting Recurrence

20 items • Updated 6 days ago • 6