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

cffffcc

-0.81

ascript

-0.81

hap

-0.77

 misunder

-0.76

hirt

-0.75

ermanent

-0.68

uscript

-0.68

urry

-0.67

Untitled

-0.67

Page

-0.67

POSITIVE LOGITS

 mill

0.71

 duct

0.70

 CONTROL

0.66

strap

0.64

 centers

0.64

 Mechdragon

0.63

izers

0.63

 center

0.62

 Piper

0.61

zel

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.