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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hedon

-0.74

ãĥĺ

-0.70

Nap

-0.68

forth

-0.66

uphem

-0.65

çľ

-0.64

ãĤ§

-0.63

çīĪ

-0.63

 Frankfurt

-0.63

 Cheong

-0.62

POSITIVE LOGITS

 Standard

1.04

 Stacy

0.73

NOW

0.66

ulet

0.65

 Miracle

0.65

 Stout

0.65

rill

0.63

acia

0.62

MIS

0.61

antha

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.