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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

OPLE

-0.85

 Penguins

-0.70

rongh

-0.70

Commission

-0.69

Pes

-0.69

WER

-0.68

POS

-0.66

YP

-0.65

Coun

-0.64

cession

-0.64

POSITIVE LOGITS

ima

0.73

lets

0.70

ides

0.68

ont

0.67

ama

0.67

mins

0.66

ane

0.66

ysis

0.66

oga

0.65

ibu

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.