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

âĹ¼

-0.80

 quotation

-0.68

Je

-0.66

 sear

-0.64

Phones

-0.62

Esk

-0.62

 caption

-0.61

 amput

-0.61

hig

-0.61

 typew

-0.61

POSITIVE LOGITS

ira

0.94

ageddon

0.79

onent

0.76

irin

0.76

aeda

0.75

erva

0.74

 sidx

0.72

omever

0.72

atari

0.71

aiden

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.