INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bs

-0.82

IP

-0.71

angular

-0.71

Software

-0.70

à¦

-0.69

bsite

-0.69

Micro

-0.68

enough

-0.68

Original

-0.67

Serial

-0.66

POSITIVE LOGITS

uba

0.74

uary

0.67

 Sabb

0.67

cking

0.65

hower

0.65

ounge

0.63

 whine

0.62

avement

0.62

aneously

0.61

Ã°

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.