VISTA: Steering LLMs with Latent User Values Inferred from a Few Survey Answers

Paper recorded by Signals 4 on 2026-10-16 in cs.CL. Abstract reproduced from arXiv; link to the original below.

Published 2026-10-16 on arXiv · recorded by Signals 4 on 2026-10-19

Category: cs.CL · 自然语言处理 · first seen 2026-10-19

Abstract

Personalizing a language model requires knowing what a person values, which stays largely consistent across tasks. Yet most methods either learn task-specific preferences or place the user's profile in the prompt, where models struggle to act on it. We introduce VISTA (Value-Informed Steering and Targeted Acquisition), which infers a latent profile from users' (full or partial) survey answers, and

Read on arXiv →