LearnLM: Improving Gemini for Learning

TL;DR

LearnLM introduces a pedagogical instruction following approach to enhance Gemini models for educational purposes, enabling customizable teaching behaviors and significantly improving expert preferences across diverse learning scenarios.

Contribution

The paper presents a novel training framework for Gemini models using pedagogical instruction following, allowing customizable teaching behaviors and improved learning performance.

Findings

LearnLM models are substantially preferred by experts over GPT-4o, Claude 3.5 Sonnet, and Gemini 1.5 Pro.

Training with pedagogical instructions improves model suitability for educational tasks.

The approach enables flexible integration of pedagogical data into existing models.

Abstract

Today's generative AI systems are tuned to present information by default, rather than engage users in service of learning as a human tutor would. To address the wide range of potential education use cases for these systems, we reframe the challenge of injecting pedagogical behavior as one of \textit{pedagogical instruction following}, where training and evaluation examples include system-level instructions describing the specific pedagogy attributes present or desired in subsequent model turns. This framing avoids committing our models to any particular definition of pedagogy, and instead allows teachers or developers to specify desired model behavior. It also clears a path to improving Gemini models for learning -- by enabling the addition of our pedagogical data to post-training mixtures -- alongside their rapidly expanding set of capabilities. Both represent important changes from our initial tech report. We show how training with pedagogical instruction following produces a LearnLM model (available on Google AI Studio) that experts substantially prefer across a diverse set of learning scenarios, with average preference strengths of +31\% over GPT-4o, +11\% over Claude 3.5 Sonnet, and +13\% over the Gemini 1.5 Pro model on which LearnLM was based.