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

gins

-0.82

baugh

-0.78

 Rudolph

-0.66

eson

-0.66

Rex

-0.65

rely

-0.65

 millenn

-0.64

ger

-0.61

tyr

-0.61

Engineers

-0.61

POSITIVE LOGITS

bably

0.68

odcast

0.63

asca

0.62

igne

0.62

arbon

0.62

 raided

0.61

WD

0.61

utic

0.60

 <!--

0.60

DRM

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.