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

PA

-0.82

Rush

-0.79

ARA

-0.73

rador

-0.70

enza

-0.67

Wiki

-0.66

pa

-0.65

rian

-0.64

VA

-0.63

uesday

-0.62

POSITIVE LOGITS

\\\\

0.76

 Harlem

0.68

 Hamburg

0.67

ãĥİ

0.67

amus

0.66

querade

0.66

EngineDebug

0.65

 Shutterstock

0.65

utterstock

0.64

lled

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.