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

ithing

-0.77

efully

-0.75

CV

-0.72

 Prior

-0.67

<<

-0.66

edly

-0.63

uphem

-0.62

yond

-0.62

fal

-0.61

gey

-0.61

POSITIVE LOGITS

rance

0.73

ECT

0.69

auga

0.62

Mesh

0.62

 Kuala

0.61

 lungs

0.60

ulum

0.60

 pals

0.60

MEN

0.60

 Saud

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.