Summary of Foundation Model Transparency Reports, by Rishi Bommasani et al.
Foundation Model Transparency Reports
by Rishi Bommasani, Kevin Klyman, Shayne Longpre, Betty Xiong, Sayash Kapoor, Nestor Maslej, Arvind Narayanan, Percy Liang
First submitted to arxiv on: 26 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
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 Foundation models have significant societal impact, necessitating transparency about their development and deployment. To establish standards for this transparency, we propose Foundation Model Transparency Reports, drawing on social media’s transparency reporting practices. Our goal is to institutionalize transparency reporting before the industry matures. We identified six design principles based on successful and unsuccessful aspects of social media transparency reporting. Additionally, we drew upon the 100 transparency indicators from the Foundation Model Transparency Index. By aligning these indicators with regulatory requirements from prominent government policies (e.g., EU AI Act, US Executive Order), well-designed reports could reduce compliance costs due to overlapping requirements across jurisdictions. We recommend that foundation model developers regularly publish transparency reports, building on recommendations from the G7 and White House. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Foundation models have a big impact on society. To make sure we understand how they’re made and used, we need transparent reporting. Imagine getting a report card for your favorite AI model! That’s what our proposal is all about – creating a standard way to share information about foundation models. We looked at how social media companies share their secrets and applied those lessons to create guidelines for transparency reports. We also drew inspiration from an index of 100 things that should be reported transparently when it comes to AI. By matching these indicators with rules from government policies, we can make sure companies are being honest and don’t have to repeat the same effort in different places. This way, they can save time and money. |
