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

reau

-0.97

amera

-0.85

acci

-0.83

 glac

-0.75

illon

-0.69

eatures

-0.68

geon

-0.68

galitarian

-0.67

zers

-0.67

 regenerate

-0.66

POSITIVE LOGITS

sil

0.81

MU

0.75

Gaming

0.72

Sil

0.70

advertising

0.69

MU

0.69

British

0.67

lees

0.67

 Petition

0.66

Â£

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.