Guide Labs

@guidelabsai

Engineering interpretable AI systems that are easy to understand, trust, and debug.

San Francisco, CA

Joined January 2023

Pinned
Guide Labs
@guidelabsai
Jun 11
Today we’re announcing a finding that breaks a core assumption in AI: that bigger models are harder to understand. We show the opposite. When interpretability is built into training, models become MORE understandable as they become more capable.
00:00
Guide Labs
@guidelabsai
Aug 13
UPDATE: Steerling-8b has broken through the assumption that larger models + more data are harder to interpret → Interpretability can be forecasted before an expensive run, and we can estimate ROI from params and data. Instead of a black box you have to reverse-engineer
Guide Labs
@guidelabsai
Jul 2
Typical training pipeline: tokens + concepts → model Clarity: model → concepts → steerability
Guide Labs
@guidelabsai
Jul 2
Every word this model generates is driven by concepts you can see...and change. This is Clarity: the first inherently interpretable LLM
00:00
Guide Labs
@guidelabsai
Jun 30
Interpretability is a fixed cost. We scaled inherently interpretable LLMs from 10M→8B params. They only became smarter and easier to understand.