INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

east

-0.78

alyst

-0.77

QL

-0.76

rounder

-0.75

paralle

-0.74

tip

-0.74

 grocer

-0.74

ioch

-0.71

ilingual

-0.68

shift

-0.68

POSITIVE LOGITS

 Metall

0.71

doi

0.66

 reversed

0.63

 watered

0.61

iss

0.61

amus

0.60

 variance

0.58

 interchangeable

0.58

 Mechdragon

0.58

 Babel

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.