[Talk] Surveying AI Safety Research Directions

We are a community of people interested in ensuring that the development of artificial intelligence benefits humanity’s long-term future. We are based in Edinburgh, but also act as a hub for surrounding areas.

AI Safety Hub Edinburgh

Dan Hendrycks gives a whirlwind tour of directions in AI safety: withstanding hazards, identifying hazards, and reducing ML system hazards

<h1>Surveying AI Safety Research Directions</h1><p>Join AI Safety Hub Edinburgh (<a href="https://aished.org" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">AISHED</a>) for a talk by Dan Hendrycks surveying different research directions in AI safety.</p><p>Attendance is free. If you&#x27;re interested in being informed of more talks and events run by AISHED, <a href="https://aished.org" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">visit our website</a> to sign up to our <a href="https://mailchi.mp/9f9a3197ec71/mailing-list-signup" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">mailing list</a>. </p><h2>Talk Abstract</h2><p>ML systems are rapidly increasing in size, are acquiring new capabilities, and are increasingly deployed in high-stakes settings.  In this presentation, I’ll give a whirlwind tour of directions in safety, namely withstanding hazards (“Robustness”), identifying hazards (“Monitoring”), reducing inherent ML system hazards (“Alignment”).</p><h2>Speaker Bio</h2><p><a href="https://people.eecs.berkeley.edu/~hendrycks/" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">Dan Hendrycks</a> received his PhD from UC Berkeley where he was advised by Dawn Song and Jacob Steinhardt. He is now the director of the <a href="https://safe.ai/" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">Center for AI Safety</a>. Dan&#x27;s research is supported by the NSF GRFP and the Open Philanthropy AI Fellowship. Dan helped contribute the <a href="https://paperswithcode.com/method/gelu" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">GELU activation function</a> (the most-used activation in state-of-the-art models including BERT, GPT, Vision Transformers, etc.), the <a href="https://arxiv.org/abs/1610.02136" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">out-of-distribution detection baseline</a>, and <a href="https://arxiv.org/abs/1903.12261" target="_blank" data-msys-clicktrack="0" rel="nofollow noopener noreferrer">distribution shift benchmarks</a>.</p>

Our {communityGuidelinesLink} describe the sort of content we prohibit on Eventbrite. If you suspect an event may be in violation, you can report it to us so we can investigate.

To report other categories of prohibited or illegal content, submit {link}

[Talk] Surveying AI Safety Research Directions

More events from AI Safety Hub Edinburgh

Discover more events from AI Safety Hub Edinburgh, from Science & Tech to other experiences you might love.

Still looking for the right event?

Explore all online events to browse and filter by date, category, and more.

[Talk] Surveying AI Safety Research Directions

More events from AI Safety Hub Edinburgh

Discover more events from AI Safety Hub Edinburgh, from Science & Tech to other experiences you might love.

You might also like...

Browse more events with different dates, prices, and formats to find your next great experience.

Still looking for the right event?

Explore all online events to browse and filter by date, category, and more.