Language Models Struggle to Use Representations Learned In-Context

TL;DR

This paper investigates whether large language models can effectively use in-context learned representations for downstream tasks, revealing significant struggles even in state-of-the-art models, and aims to inspire new methods for better deployment of in-context information.

Contribution

It demonstrates that current open- and closed-source LLMs struggle to utilize in-context representations for task completion, highlighting a gap in flexible deployment of learned in-context information.

Findings

Open-weights LLMs struggle with in-context representations for next-token prediction.

State-of-the-art reasoning models cannot reliably leverage novel in-context patterns.

Models encode in-context information but fail to deploy it effectively for downstream tasks.

Abstract

Though large language models (LLMs) have enabled great success across a wide variety of tasks, they still appear to fall short of one of the loftier goals of artificial intelligence research: creating an artificial system that can adapt its behavior to radically new contexts upon deployment. One important step towards this goal is to create systems that can induce rich representations of data that are seen in-context, and then flexibly deploy these representations to accomplish goals. Recently, Park et al. (2024) demonstrated that current LLMs are indeed capable of inducing such representation from context (i.e., in-context representation learning). The present study investigates whether LLMs can use these representations to complete simple downstream tasks. We first assess whether open-weights LLMs can use in-context representations for next-token prediction, and then probe models using a novel task, adaptive world modeling. In both tasks, we find evidence that open-weights LLMs struggle to deploy representations of novel semantics that are defined in-context, even if they encode these semantics in their latent representations. Furthermore, we assess closed-source, state-of-the-art reasoning models on the adaptive world modeling task, demonstrating that even the most performant LLMs cannot reliably leverage novel patterns presented in-context. Overall, this work seeks to inspire novel methods for encouraging models to not only encode information presented in-context, but to do so in a manner that supports flexible deployment of this information.