en
每月不到10元,就可以无限制地访问最好的AIbase。立即成为会员
Home
News
Daily Brief
Income Guide
Tutorial
Tools Directory
Product Library
en
Search AI Products and News
Explore worldwide AI information, discover new AI opportunities
AI News
AI Tools
AI Cases
AI Tutorial
Type :
AI News
AI Tools
AI Cases
AI Tutorial
2024-08-01 10:31:36
.
AIbase
.
10.7k
Unlocking the 'Black Box' of Language Models! Google DeepMind Releases a Visualization Tool Gemma Scope
Google DeepMind's latest research, Gemma Scope, unveils the secrets of the 'black box' of language models by using Sparse Autoencoders (SAEs) to decompose and reconstruct the activations of language models, aiming to reveal meaningful features. Gemma Scope employs JumpReLU SAEs, which optimize reconstruction loss and regularize the number of active latent features by controlling activations to reveal the internal mechanisms of language models. The research found that the performance of residual flow SAEs is generally low, and sequence length affects the performance of SAEs.