We use GPT-4 to automatically write explanations for the behavior of neurons in large language models and to score those explanations. We release a dataset of these (imperfect) explanations and scores for every neuron in GPT-2.
Recommended
Language models are few-shot learnersOpenai Simplifying, stabilizing, and scaling continuous-time consistency modelsOpenai Operation "Doppelganger": Russian influence activity targeting UkraineOpenai Everyday work with ChatGPT WorkOpenai Romance scams: AI-enabled romance scam workflowsOpenai Will Hurd joins OpenAI’s board of directorsOpenai
Breaking down common documentation mistakesMintlify
10 best AI observability tools for monitoring and evaluating agents in 2026Mintlify
Trained on 100,000+ Voices: Deepgram Unveils Next-Gen Speaker Diarization and Language DetectionDeepgram
The Language of LGBTQ Inclusion and Allyship - Deepgram Blog ⚡️Deepgram
Calling Your Video Game With Your Phone: Part 1Deepgram
The Best (and Worst?) of Voice AI TechnologyDeepgram
New Spanish and Turkish Language Models and Updated General Models - Deepgram Blog ⚡️Deepgram
Data Ingestion: 6 Ways to Speed Up Your ApplicationRedis
How to Conduct an AI Agent Security Audit [+ Checklist]Zapier
How to implement AI training for employeesZapier
AI frameworks: Definition, types, and how to chooseZapier
7 ways to use Zapier's Zendesk integrationZapier
How to connect Google Sheets to NotionZapier
How Argo CD and OpenShift Enable GitOps for DevelopersThenewstack
