LearnLM: Improving Gemini for Learning
LearnLM Team, Abhinit Modi, Aditya Srikanth Veerubhotla, Aliya Rysbek, Andrea Huber, Brett Wiltshire, Brian Veprek, Daniel Gillick, Daniel Kasenberg, Derek Ahmed, Irina Jurenka, James Cohan, Jennifer She, Julia Wilkowski, Kaiz Alarakyia, Kevin McKee, Lisa Wang, Markus Kunesch, Mike Schaekermann, Miruna Pîslar, Nikhil Joshi, Parsa Mahmoudieh
2024-12-24
Summary
This paper talks about LearnLM, a new AI model developed by Google that aims to improve learning by acting like a personal tutor. It uses advanced techniques to help students understand subjects better through interactive and personalized teaching methods.
What's the problem?
Many generative AI systems focus on providing information rather than engaging with users in a way that promotes learning, similar to how a human tutor would. This lack of interaction can make learning less effective and less engaging for students, especially when dealing with complex topics.
What's the solution?
To solve this problem, LearnLM introduces a method called pedagogical instruction following, which allows teachers or developers to specify how the model should behave in teaching scenarios. By training LearnLM with examples that demonstrate good teaching practices, it can provide more effective and personalized responses. The model has been tested and shown to perform better than previous models like GPT-4o and Claude 3.5 in various learning situations, making it a valuable tool for education.
Why it matters?
This research is important because it represents a significant step forward in using AI for education. By creating a model that can adapt to individual learning needs and provide interactive support, LearnLM has the potential to enhance educational experiences for students, making learning more accessible and effective across different subjects.
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 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 is preferred substantially by expert raters across a diverse set of learning scenarios, with average preference strengths of 31\% over GPT-4o, 11\% over Claude 3.5, and 13\% over the Gemini 1.5 Pro model LearnLM was based on.