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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Forsaken

-0.67

 Malf

-0.66

 Flavoring

-0.66

 horrors

-0.66

 Decay

-0.65

Redditor

-0.64

 legion

-0.64

 retribution

-0.63

 Remastered

-0.63

BILITY

-0.62

POSITIVE LOGITS

axis

0.85

enium

0.80

icycle

0.80

ione

0.80

colo

0.77

tyard

0.76

kees

0.75

ibraries

0.75

strip

0.74

ublic

0.74

Activations Density 0.000%

No Known Activations

This feature has no known activations.