Summary of Looking Into Black Box Code Language Models, by Muhammad Umair Haider et al.
Looking into Black Box Code Language Models
by Muhammad Umair Haider, Umar Farooq, A.B. Siddique, Mark Marron
First submitted to arxiv on: 5 Jul 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Software Engineering (cs.SE)
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 A recent surge in research has focused on applying Language Models (LMs) to tasks related to coding, with several code-LMs proposed to improve performance on various benchmarks. While LMs have been shown to excel in these areas, the majority of studies neglect to explore the inner workings of these models, treating them as black boxes. A few efforts have attempted to understand the role of attention layers in code-LMs, but the vast majority of parameters in a typical transformer model remain unexplored, specifically the feed-forward layers that make up two-thirds of its architecture. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Code language models have been used for coding-related tasks and many new models have been developed. These models are like black boxes, even though some people are trying to understand how attention works in them. Most researchers haven’t looked at what makes these models work, especially the feed-forward layers that make up most of a typical model. |
Keywords
» Artificial intelligence » Attention » Transformer
