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

 Shooter

-0.80

CPC

-0.71

 Beach

-0.67

 è£ıè

-0.66

 Sinai

-0.66

 Skies

-0.64

 Creek

-0.64

¿½

-0.63

 Forward

-0.63

%%

-0.61

POSITIVE LOGITS

uries

0.74

ebook

0.71

cens

0.70

inous

0.70

ocrats

0.69

ially

0.69

iership

0.69

 loaf

0.68

pees

0.68

ellery

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.