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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ollah

-0.79

asma

-0.75

ificantly

-0.73

Forge

-0.72

risome

-0.72

ially

-0.70

orb

-0.70

weet

-0.70

ule

-0.70

olls

-0.70

POSITIVE LOGITS

 dysph

0.80

 surv

0.68

 compartment

0.68

 adventurous

0.67

 alike

0.63

hun

0.62

 srfAttach

0.62

 homeless

0.61

 cruiser

0.61

Ô

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.