Words in Motion: Extracting Interpretable Control Vectors for Motion Transformers

Autor:	Tas, Omer Sahin, Wagner, Royden
Rok vydání:	2024
Předmět:	Computer Science - Machine Learning Computer Science - Computation and Language Computer Science - Computer Vision and Pattern Recognition
Druh dokumentu:	Working Paper
Popis:	Transformer-based models generate hidden states that are difficult to interpret. In this work, we aim to interpret these hidden states and control them at inference, with a focus on motion forecasting. We leverage the phenomenon of neural collapse and use linear probes to measure interpretable features in hidden states. Our experiments reveal meaningful directions and distances between hidden states of opposing features, which we use to fit control vectors for activation steering. Consequently, our method enables controlling transformer-based motion forecasting models with interpretable features, providing a unique interface to interact with and understand these models. Our implementation is available at https://github.com/kit-mrt/future-motion Comment: NeurIPS'24 workshop camera-ready
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2406.11624 Zobrazit plný text záznamu View this record from Arxiv