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

-0.90

-0.86

 Palestin

-0.76

Compat

-0.71

 Couch

-0.70

},{"

-0.68

 supra

-0.66

rug

-0.65

ERA

-0.64

POSITIVE LOGITS

inion

0.75

anos

0.75

veyard

0.73

nesses

0.71

undown

0.71

zin

0.69

stroke

0.68

imaru

0.67

igham

0.67

ete

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.