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

aimon

-0.85

natureconservancy

-0.84

 indo

-0.82

wcsstore

-0.80

ĻĤ

-0.77

é¾įå¥ĳå£«

-0.73

emet

-0.73

entimes

-0.72

ãĤ§

-0.71

unia

-0.71

POSITIVE LOGITS

tsy

0.70

 Proud

0.69

 Welsh

0.67

FT

0.64

 Liberty

0.63

 Bristol

0.61

 Wales

0.61

Ner

0.59

 Sinclair

0.59

 Lady

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.