Human-like Cognitive Generalization for Large Models via Brain-in-the-loop Supervision

TL;DR

This paper demonstrates that brain-in-the-loop supervision can transfer human conceptual understanding to large models, significantly improving their ability to reason, generalize, and interpret abstract concepts, thus making AI more human-like.

Contribution

Introducing brain-in-the-loop supervised learning to enhance large models' cognitive abilities with minimal brain signal data, improving abstract reasoning and interpretability.

Findings

Enhanced performance in few-shot and zero-shot tasks

Improved out-of-distribution recognition

More interpretable concept representations

Abstract

Recent advancements in deep neural networks (DNNs), particularly large-scale language models, have demonstrated remarkable capabilities in image and natural language understanding. Although scaling up model parameters with increasing volume of training data has progressively improved DNN capabilities, achieving complex cognitive abilities - such as understanding abstract concepts, reasoning, and adapting to novel scenarios, which are intrinsic to human cognition - remains a major challenge. In this study, we show that brain-in-the-loop supervised learning, utilizing a small set of brain signals, can effectively transfer human conceptual structures to DNNs, significantly enhancing their comprehension of abstract and even unseen concepts. Experimental results further indicate that the enhanced cognitive capabilities lead to substantial performance gains in challenging tasks, including few-shot/zero-shot learning and out-of-distribution recognition, while also yielding highly interpretable concept representations. These findings highlight that human-in-the-loop supervision can effectively augment the complex cognitive abilities of large models, offering a promising pathway toward developing more human-like cognitive abilities in artificial systems.