INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Veh

-0.68

unin

-0.65

 vying

-0.64

 distracted

-0.63

 contended

-0.60

assetsadobe

-0.60

 ownership

-0.60

 slug

-0.59

mentation

-0.58

clair

-0.58

POSITIVE LOGITS

atum

0.93

adle

0.75

idi

0.73

APH

0.70

ARP

0.67

Ã¸

0.67

 continuum

0.67

HI

0.67

 Fein

0.67

APS

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.