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

igious

-0.80

UAL

-0.80

igion

-0.78

 Bund

-0.69

ingu

-0.67

Material

-0.66

ieve

-0.65

acia

-0.64

court

-0.63

ATURE

-0.63

POSITIVE LOGITS

abouts

0.87

rums

0.74

 Moons

0.74

 suspic

0.72

 horizont

0.70

 undermin

0.70

izen

0.68

 antioxid

0.67

Þ

0.66

RS

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.