Paper · arXiv 2408.05506
Your Context Is Not an Array: Unveiling Random Access Limitations in Transformers
MohammadReza Ebrahimi, Sunny Panchal, Roland Memisevic
Transformer-based models fail to generalize to longer problem instances due to an inability to randomly access memory within their context window, as demonstrated through analysis of attention maps and examination of effective memory access methodologies.
Abstract
Despite their recent successes, Transformer-based large language models show surprising failure modes. A well-known example of such failure modes is their inability to length-generalize: solving problem instances at inference time that are longer than those seen during training. In this work, we further explore the root cause of this failure by performing a detailed analysis of model behaviors on the simple parity task. Our analysis suggests that length generalization failures are intricately related to a model's inability to perform random memory accesses within its context window. We present supporting evidence for this hypothesis by demonstrating the effectiveness of methodologies that circumvent the need for indexing or that enable random token access indirectly, through content-based addressing. We further show where and how the failure to perform random memory access manifests through attention map visualizations.