From Poetry to Programming: The Evolution of Prompt Engineering with Riley Goodside of Scale AI
Nathan hosts Riley Goodside, the world's first staff prompt engineer at Scale AI, to discuss the evolution of
Explore the frontier of AI interpretability with Arthur Conmy's ACDC approach, automating the discovery of sub-circuits within transformers. Sponsored by NetSuite.
Join Arthur Conmy and Nathan Labenz in this captivating and accessible discussion as they embark on a deep dive into the cutting-edge world of interpretability research. Discover how pioneering researchers have isolated sub-circuits within transformers that are responsible for different aspects of AI capacity. Arthur introduces us to the groundbreaking ACDC approach, a revolutionary method co-authored by him, which automates the most time-consuming aspects of this intricate research. If you’re looking for an ERP platform, check out our sponsor, NetSuite: http://netsuite.com/cognitive
TIMESTAMPS:
(00:00) Episode Preview
(04:40) What attracted Arthur to mechanistic interpretability?
(07:49) LLM information processing: General Understanding vs Stochastic Parrot Paradigm
(14:00) ACDC paper: https://arxiv.org/abs/2304.14997
(14:45) Sponsors: NetSuite | Omneky
(24:30) Putting together data sets
(32:39) How to intervene in LLMs network activity
(36:00) Setting metrics to evaluate the production of correct completions
(44:20) The future of the mechanistic interpretability research
(50:00) Extracting upstream activations in the ACDC project and evaluating impact on downstream components.
(56:00) Anthropic research findings
(01:08:00) 3-Step process of the ACDC approach
(01:22:00) Setting a threshold and validation
(01:27:00) Goal of the approach
(01:32:00) Compute requirements
*Correction at 1:33:00 Arthur meant to say = "quadratic in nodes"
(01:35:30) Scaling laws for mechanistic interpretability
(01:40:00) Accessibility of this research for casual enthusiasts
(01:46:00) Emergence discourse
(01:56:00) Path to AI safety
LINKS:
Towards Automated Circuit Discovery for Mechanistic Interpretability https://arxiv.org/abs/2304.14997
https://arthurconmy.github.io/
SOCIAL MEDIA:
@labenz (Nathan)
@arthurconmy (Arthur)
@cogrev_podcast
SPONSORS: NetSuite | Omneky
-NetSuite provides financial software for all your business needs. More than thirty-six thousand companies have already upgraded to NetSuite, gaining visibility and control over their financials, inventory, HR, eCommerce, and more. If you’re looking for an ERP platform: NetSuite (Code http://netsuite.com/cognitive) and defer payments of a FULL NetSuite implementation for six months.
-Omneky is an omnichannel creative generation platform that lets you launch hundreds of thousands of ad iterations that *actually work* customized across all platforms, with a click of a button. Omneky combines generative AI and real-time advertising data. Mention "Cog Rev" for 10% off.
Be the first to know when new episodes drop. Subscribe now for exclusive insights from AI trailblazers and in-depth analysis straight to your inbox.