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

PDATE

-0.68

comm

-0.68

 Leban

-0.67

 Archdemon

-0.67

yp

-0.65

doms

-0.65

tyard

-0.63

 Rebell

-0.62

+(

-0.62

config

-0.62

POSITIVE LOGITS

ocular

0.68

 goto

0.66

 Pearce

0.63

Â¢

0.61

ilts

0.61

à©

0.61

 Santiago

0.60

 discussed

0.60

 Helpful

0.59

 quoted

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.