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

Enlarge

-0.84

ester

-0.76

 Sweeney

-0.75

ebus

-0.73

esters

-0.71

stall

-0.69

NPR

-0.68

.</

-0.66

thro

-0.65

APD

-0.64

POSITIVE LOGITS

 proble

0.70

ront

0.69

 stud

0.67

 redes

0.66

 DRAG

0.65

 plent

0.65

hower

0.63

 Creatures

0.63

 misunder

0.63

 mathemat

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.