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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uss

-0.72

cn

-0.72

phabet

-0.67

rem

-0.64

});

-0.63

yll

-0.61

zz

-0.61

 Bran

-0.60

ornia

-0.60

anie

-0.60

POSITIVE LOGITS

hops

0.69

Ĭ±

0.68

Pinterest

0.67

 epigen

0.65

chrom

0.62

meier

0.62

AIDS

0.61

 Apex

0.60

henko

0.60

HUD

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.