Living comparison table of LLM architectural choices (norm, attention, MoE, positional encoding, and more) from the Original Transformer (2017) to frontier models (2026). Based on Harm de Vries's f…

Homepage of Yichen (Zach) Wang.

This is the code for BACo proposed in Optimizing Diversity and Quality through Base–Aligned Model Collaboration.

Papers and resources about LLMs creavitity in open-ended generation.

A curated collection of research papers exploring diversity in Large Language Model text generation. This repository tracks cutting-edge research on methods, evaluation, annalysis, and application.…

An automated maintenance tool for curated academic awesome paper lists on Github. Simply maintain a plain text file containing paper URLs. The system will automatically handles formatting, sectioni…

This repository includes the code implementation of the paper Improving Pacing in Long-Form Story Planning by Yichen Wang, Kevin Yang, Xiaoming Liu, and Dan Klein.

The code implementation of the paper Stumbling Blocks: Stress Testing the Robustness of Machine-Generated Text Detectors Under Attacks (ACL 2024 main) by Yichen Wang, Shangbin Feng, Abe Bohan Hou, …

The code implementation of the paper CoCo: Coherence-Enhanced Machine-Generated Text Detection Under Low Resource With Contrastive Learning from XJTU.

Config files for my GitHub profile.