Author

Ziwei Gu

Institution: Harvard University

Observed institution: Harvard University

OpenAlex ID: A5148153671

Sampling Reveals Style: Unsupervised, Training-Free Discovery of Prompt-Conditional Stylistic Axes in LLM Activations

2026 · Ajit Mallavarapu, Ziwei Gu · BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP (BlackboxNLP) · 1

Works count: 1

Citation count: 1

H-index: 1

I10-index: