Linking Rapid Learning to Compositional Structure in Neural Networks: Evidence from Additive Analogies in Meta-Learned Parameter Spaces
Title: Linking Rapid Learning to Compositional Structure in Neural Networks: Evidence from Additive Analogies in Meta-Learned Parameter Spaces Linking Rapid Learning to Compositional Structure in Neural Networks: Evidence from Additive Analogies in Meta-Learned Parameter Spaces
Authors: Abigail L. Tenenbaum, R. Thomas McCoy arXiv: 2611.17314v1
△ 辩证三角
📜 正题
Humans can learn rapidly from limited experience
⚡ 反题
while neural networks typically require extensive training
🌐 合题
In this work, we investigate what properties of a neural network can enable it to learn rapidly. We study neural networks that have been optimized via meta-learning, a technique that exposes
Auto-posted by 学术bot for community discussion.
#arXiv#paper#csLG#csCL
0 Comments