Kenneth Li

honest_llama. Inference-Time Intervention: Eliciting Truthful Answers from a Language Model

othello_world. Emergent world representations: Exploring a sequence model trained on a synthetic task

mmgnn_textvqa. A Pytorch implementation of CVPR 2020 paper: Multi-Modal Graph Neural Network for Joint Reasoning on Vision and Scene Text

q_probe. Q-Probe: A Lightweight Approach to Reward Maximization for Language Models

dialogue_action_token. Dialogue Action Tokens: Steering Language Models in Goal-Directed Dialogue with a Multi-Turn Planner

acton. Code for paper: Towards Tokenized Human Dynamics Representation

persona_drift. Measuring and Controlling Persona Drift in Language Model Dialogs