NLP

Adaptive Decoding via Latent Preference Optimization

November 19, 2024

Abstract

During language model decoding, it is known that using higher temperature sampling gives more creative responses, while lower temperatures are more factually accurate. However, such models are commonly applied to general instruction following, which involves both creative and fact seeking tasks, using a single fixed temperature across all examples and tokens. In this work, we introduce Adaptive Decoding, a layer added to the model to select the sampling temperature dynamically at inference time, at either the token or example level, in order to optimize performance. To learn its parameters we introduce Latent Preference Optimization (LPO) a general approach to train discrete latent variables such as choices of temperature. Our method outperforms all fixed decoding temperatures across a range of tasks that require different temperatures, including UltraFeedback, Creative Story Writing, and GSM8K.

Download the Paper

AUTHORS

Written by

Shehzaad Dhuliawala

Ilia Kulikov

Ping Yu

Asli Celikyilmaz

Jason Weston

Sainbayar Sukhbaatar

Jack Lanchantin

Publisher

arXiv

Related Publications

July 02, 2025

REINFORCEMENT LEARNING

NLP

ASTRO: Teaching Language Models to Reason by Reflecting and Backtracking In-Context

Joongwon (Daniel) Kim, Anirudh Goyal, Liang Tan, Hannaneh Hajishirzi, Srini Iyer, Tianlu Wang

July 02, 2025

May 14, 2025

HUMAN & MACHINE INTELLIGENCE

SPEECH & AUDIO

Emergence of Language in the Developing Brain

Linnea Evanson, Christine Bulteau, Mathilde Chipaux, Georg Dorfmüller, Sarah Ferrand-Sorbets, Emmanuel Raffo, Sarah Rosenberg, Pierre Bourdillon, Jean Remi King

May 14, 2025

April 25, 2025

RESEARCH

NLP

ReasonIR: Training Retrievers for Reasoning Tasks

Rulin Shao, Qiao Rui, Varsha Kishore, Niklas Muennighoff, Victoria Lin, Daniela Rus, Bryan Kian Hsiang Low, Sewon Min, Scott Yih, Pang Wei Koh, Luke Zettlemoyer

April 25, 2025

April 17, 2025

HUMAN & MACHINE INTELLIGENCE

CONVERSATIONAL AI

Collaborative Reasoner: Self-improving Social Agents with Synthetic Conversations

Ansong Ni, Ruta Desai, Yang Li, Xinjie Lei, Dong Wang, Ramya Raghavendra, Gargi Ghosh, Daniel Li (FAIR), Asli Celikyilmaz

April 17, 2025

Help Us Pioneer The Future of AI

We share our open source frameworks, tools, libraries, and models for everything from research exploration to large-scale production deployment.