Posted on November 7, 2019
by Noon van der Silk
| Back to recent posts
Explainable AI Event - Slides and Debrief
Can humans explain their decisions via attention?
Last night we held a discussion group, here at OC House around Explainable AI. Here’s an overview of what we covered:
- What is an explanation? What kinds of questions would we like to ask a model?
- What is happening out in the research world? What are the different approaches?
- What are the alternatives to trying to “explain” deep models?
- What are some of the tools, and how can I use them?
- What does it even mean to “know” things?
- What are people out in industry using explainable AI for? (Answer: It’s not for end-users to understand models.)
Here are the slides:
- Noon - Explainable AI
- Laura - How to XAI
Thanks to those who came along! It was a really enjoyable and lively discussion!