Introducing the Model Openness Framework: Achieving...
summary
Model Openness Framework defines a practical checklist for deep‑learning models with 17 components (code, data, documentation, weights), recommends treating model weights under data licenses rather than software licenses, and uses a three‑tier classification (class 1–3 up to “open science”) to score completeness and openness. A companion Model Openness Tool runs evaluations, emits badges and JSON you can drop into GitHub/Hugging Face, and flags invalid licenses; limitations include focus on deep learning, difficulty sharing datasets, and no direct bias/safety fixes.
AI-written summary · about
chapters (6)
- 0:00Model Openness Framework intro
- 3:58Defining AI openness challenges
- 9:01Three-tiered Model Openness
- 14:38Demo of Openness Tool
- 20:41Exploring Openness Tool interface
- 25:01Call to action & community
related talks
34mThe Model Openness Framework: Promoting Completeness and Openness for Reproducibil...
29mFostering Responsible AI:Empowering Openness and Community Collaboration
37mIs Your AI Truly Open ? Cutting Through the Hype
22mThe Role of Openness in Responsible AI
32mOperationalizing Openness: Standardizing AI Model Supply Chains With the Model Openness Framework
40mThe Generative AI Commons: Unlocking Potential Through Collaboration and Transpare...
42mThe Open Source AI Dilemma: Crafting a Clear Definition for Open Source AI
41mThe Generative AI Commons: Unlocking Potential Through Collaboration and Transpare...