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

 inval

-0.78

olkien

-0.70

bons

-0.63

 Creator

-0.63

 Named

-0.60

India

-0.60

lisher

-0.58

oire

-0.58

ounge

-0.58

obook

-0.58

POSITIVE LOGITS

rawdownloadcloneembedreportprint

0.77

ħĭ

0.72

¿½

0.66

atom

0.62

xp

0.62

vous

0.61

worldly

0.61

 teen

0.60

 slowed

0.58

paced

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.