INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ouk

-0.77

doi

-0.77

¿

-0.76

yss

-0.71

leon

-0.71

ĸļ

-0.71

orno

-0.70

hani

-0.70

azaki

-0.69

gdala

-0.68

POSITIVE LOGITS

naires

0.71

tics

0.71

 Icar

0.66

 uniform

0.63

ibles

0.61

ific

0.61

LESS

0.59

ifications

0.57

 Uran

0.56

 conform

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.