Into the Unknown: Self-Learning Large Language Models
This work addresses the challenge of what to learn for self-learning LLMs, offering a method to improve model updates and knowledge acquisition, though it appears incremental as it builds on existing self-learning concepts.
The authors tackled the problem of enabling large language models (LLMs) to independently learn unknown knowledge by proposing a self-learning framework that identifies and focuses on atomic knowledge gaps, called Points in the Unknown (PiU). Their experiments showed that LLMs with at least 3B parameters and instruction training can perform self-learning effectively, leading to more efficient updates and new perspectives for knowledge exchange.
We address the main problem of self-learning LLM: the question of what to learn. We propose a self-learning LLM framework that enables an LLM to independently learn previously unknown knowledge through self-assessment of their own hallucinations. We introduce a concept called Point in the Unknown (PiU) to identify atomic knowledge unknown to a model, along with four methods for automatic PiUs identification, facilitating the creation of a self-learning loop that focuses exclusively on the absorption of currently unknown knowledge into the model. Additionally, we developed evaluation metrics to gauge an LLM's self-learning capability. Our experiments revealed that LLMs with at least 3B parameters that have undergone some instruction training would be able to perform self-learning well. We further proved the effectiveness of self-learning by comparing the performance of a model that has undergone self-learning to a model that has not. Our self-learning concept allows more efficient LLM updates and opens new perspectives for LLM knowledge exchange.