Summary of Transformerfam: Feedback Attention Is Working Memory, by Dongseong Hwang et al.
TransformerFAM: Feedback attention is working memory
by Dongseong Hwang, Weiran Wang, Zhuoyuan Huo, Khe Chai Sim, Pedro Moreno Mengibar
First submitted to arxiv on: 14 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This novel Transformer architecture, Feedback Attention Memory (FAM), addresses the limitation of quadratic attention complexity in processing infinitely long inputs. By introducing a feedback loop, FAM enables the network to attend to its own latent representations, fostering working memory and allowing indefinite sequence processing. This design seamlessly integrates with pre-trained models, showcasing improved performance on long-context tasks across various model sizes. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary In simple terms, this paper creates a new way for artificial intelligence networks called Transformers to understand very long texts or sequences of information. The issue is that these networks can only handle short sequences because it takes too much time and memory to process longer ones. The solution proposed in this paper allows the network to look back at its own thoughts and remember important details, making it possible to process sequences as long as needed. |
Keywords
» Artificial intelligence » Attention » Transformer
