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

 sexes

-0.72

 Siberian

-0.66

 Militia

-0.65

 Lizard

-0.65

 Pole

-0.65

pole

-0.63

 genders

-0.63

 manned

-0.63

 territorial

-0.63

 docks

-0.63

POSITIVE LOGITS

hal

0.81

artz

0.80

ONSORED

0.74

ARP

0.74

auer

0.70

etitive

0.69

RON

0.69

join

0.68

0.67

æ³

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.