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

Cum

-0.78

ovember

-0.75

rd

-0.71

eting

-0.70

 Spur

-0.68

 scrimmage

-0.65

 largeDownload

-0.64

uminati

-0.62

CJ

-0.62

JJ

-0.62

POSITIVE LOGITS

ľ

2.02

ãĤ¶

0.89

®

0.88

ļ

0.82

>[

0.79

ĨĴ

0.77

ĺ

0.73

brew

0.72

Ľ

0.72

ķ

0.72

Activations Density 0.000%

No Known Activations

This feature has no known activations.