INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

--+

-0.76

MRI

-0.73

 Fitness

-0.70

osterone

-0.69

buquerque

-0.69

gdala

-0.69

cookie

-0.67

*/(

-0.65

ounty

-0.65

Psy

-0.64

POSITIVE LOGITS

nered

0.70

inqu

0.62

oused

0.62

lev

0.61

ahime

0.60

 looting

0.60

 Mour

0.59

ointed

0.58

ities

0.58

 tolerated

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.