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

ICAN

-0.68

 Bened

-0.67

 Lisbon

-0.67

 Lima

-0.66

 Angola

-0.66

 Rodney

-0.66

 whip

-0.66

 Mandela

-0.65

 Nelson

-0.65

 Cinnamon

-0.63

POSITIVE LOGITS

cedented

0.85

reditary

0.84

paralleled

0.77

htaking

0.77

lez

0.77

etimes

0.75

farious

0.73

lishing

0.73

ynski

0.72

irming

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.