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

dain

-0.79

 enhancement

-0.72

 streng

-0.71

etting

-0.71

rise

-0.69

 revived

-0.68

ĸļ

-0.66

isine

-0.65

ntil

-0.65

 preceded

-0.65

POSITIVE LOGITS

ouf

0.83

Generic

0.72

 Rite

0.71

"},

0.71

Merit

0.68

 Loans

0.68

 guiIcon

0.67

AAA

0.66

ENCY

0.66

 Generic

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.