Daphne Ippolito

Command-V: Training-Free Representation Finetuning Transfer

FictionalQA: A Dataset for Studying Memorization and Knowledge Acquisition

On Code-Induced Reasoning in LLMs

Human-Aligned Chess With a Bit of Search

Measuring Non-Adversarial Reproduction of Training Data in Large Language Models

Scalable Extraction of Training Data from Aligned, Production Language Models

NoveltyBench: Evaluating Language Models for Humanlike Diversity

Exploring and Mitigating Adversarial Manipulation of Voting-Based Leaderboards

Persistent Pre-training Poisoning of LLMs