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

itude

-0.65

 Javascript

-0.64

 Hover

-0.60

yles

-0.58

igne

-0.56

igel

-0.56

 Expand

-0.55

 warrant

-0.55

ista

-0.54

 Cyprus

-0.54

POSITIVE LOGITS

hol

1.56

washed

0.67

idated

0.67

 Hebdo

0.67

 Breaker

0.66

fighter

0.66

steen

0.66

externalActionCode

0.65

Accessory

0.65

Pub

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.