AI BriefingAI21 Achieves Top Score in Stanford University's Transparency Index
Key point
AI21 Labs recorded the highest score among commercial LLMs in Stanford University's Foundation Model Transparency Index (FMTI).
Details
AI21 Labs recorded the highest score among commercial general-purpose LLMs in Stanford University's Foundation Model Transparency Index (FMTI), also ranking 2nd overall. AI21's score surged from 25 points in October 2023 to 75 points in just 6 months, showing the largest improvement among the companies surveyed.
In this result, AI21 outperformed major competitors:
- AI21: 75 points
- Meta: 60 points
- Anthropic: 51 points
- OpenAI: 49 points
AI21 received high marks across various indicators, including data sourcing and augmentation methods, compute and hardware specifications, technical characteristics of the Jurassic-2 model, and usage policies and customer feedback mechanisms. The FMTI is led by Stanford CRFM and HAI, and evaluates the transparency of LLM developers through 13 indicators and 100 detailed items covering data, compute, and methodology.
AI21 is also focused on building reliable models for heavily regulated industries such as finance and healthcare. In particular, the Contextual Answers model, which generates answers based on an organization's internal knowledge base, effectively prevents hallucinations by grounding its answers in the provided context.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.