Lattice AI Research

Research focus

Data Curation & Synthetic Data (2)Multimodal Models (2)Eval Frameworks & Benchmarks (2)Constitutional AI & AI Ethics (1)

Frequent co-authors

Shubhashis Nobin Sarwar (1)Roy Dipta (1)Zheyuan Liu (1)Vaidehi Patil (1)

Papers (3)

Jul 8, 2026

Tsinghua AIJul 8, 2026·also AI2, Amazon Science, CMU ML, MIT CSAIL +9

Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks

Multimodal unlearning could revolutionize how we handle sensitive data in AI, enabling targeted removal without sacrificing model performance.

Shubhashis Nobin Sarwar, Roy Dipta, Zheyuan Liu +165

Constitutional AI & AI Ethics Data Curation & Synthetic Data Multimodal Models

Apr 24, 2026

Chengye Wang +3Apr 24, 2026

TexOCR: Advancing Document OCR Models for Compilable Page-to-LaTeX Reconstruction

Existing document OCR models fail to preserve crucial structural and executable properties of LaTeX, but TexOCR, trained with verifiable rewards, finally delivers compilable page-to-LaTeX reconstruction.

Chengye Wang, Ling Fu, Zexi Kuang +1

Code Generation & Program Synthesis Computer Vision Eval Frameworks & Benchmarks

Mar 12, 2026

AI2Mar 12, 2026·also Yale

SciMDR: Benchmarking and Advancing Scientific Multimodal Document Reasoning

Training on SciMDR, a new 300K QA dataset synthesized from scientific papers, substantially boosts model performance on complex, document-level scientific reasoning tasks.

Chengye Wang, Rilyn Han, Manasi S. Patwardhan +2

Data Curation & Synthetic Data Eval Frameworks & Benchmarks Multimodal Models

Search

Chengye Wang

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (3)