5 papers
Data-Constrained Language Model Pretraining: Improved Regularization and Scaling Laws
Zhiwei Xu, Shihao Wu, Hanseul Cho +2
Classical scaling laws for language model pretraining balance model size against training dataset size under a fixed compute budget, assuming abundant data and a single pass over t…
Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces
Mike A. Merrill, Alexander G. Shaw, Nicholas Carlini +82
AI agents may soon become capable of autonomously completing valuable, long-horizon tasks in diverse domains. Current benchmarks either do not measure real-world tasks, or are not…
Inference of Dependency Knowledge Graph for Electronic Health Records
Zhiwei Xu, Ziming Gan, Doudou Zhou +3
The effective analysis of high-dimensional Electronic Health Record (EHR) data, with substantial potential for healthcare research, presents notable methodological challenges. Empl…
Let Me Grok for You: Accelerating Grokking via Embedding Transfer from a Weaker Model
Zhiwei Xu, Zhiyu Ni, Yixin Wang +1
''Grokking'' is a phenomenon where a neural network first memorizes training data and generalizes poorly, but then suddenly transitions to near-perfect generalization after prolong…
Benign Overfitting in Single-Head Attention
Roey Magen, Shuning Shang, Zhiwei Xu +3
The phenomenon of benign overfitting, where a trained neural network perfectly fits noisy training data but still achieves near-optimal test performance, has been extensively studi…