Authors

Shanbin Yu

Publications 2

Preprint Open access

Forking: Sudden Overfitting Under Replay

Shanbin Yu, Shaoyang Guo, Haoran Zhao et al. · 2026

This paper studies forking, a generalization failure discovered in NanoGPT autoresearch. Under data replay, models with an over-encoding n-gram memory branch show a sharp separation of training and validation loss at epoch boundaries, resembling the shape of forks. We study this phenomenon in a controlled vanilla NanoG …

Preprint Open access

ArchitectureIQ: On the Measure of Training Intuition

Zirui Ren, Shaoyang Guo, Chencheng Tang et al. · 2026

Top researchers have good intuition, but do language models have as good intuition about model training as top AI researchers? To measure model intuition of LLMs and humans, we introduce the ArchitectureIQ benchmark. Each question presents a synthetic dataset and several training recipes, and the test-taker is asked to …

Co-authors