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

Tags

-0.72

hoff

-0.66

Hash

-0.66

atana

-0.63

avi

-0.63

 Ange

-0.62

Offline

-0.61

ä½¿

-0.61

possibly

-0.60

follow

-0.60

POSITIVE LOGITS

 rall

0.97

culosis

0.76

otin

0.75

ividual

0.73

othal

0.71

raltar

0.70

ioch

0.69

urers

0.68

lington

0.67

 manufact

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.