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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

clud

-0.83

 differe

-0.82

 consecut

-0.75

 indu

-0.72

 disag

-0.68

 implant

-0.67

inet

-0.65

hig

-0.65

inyl

-0.64

waves

-0.63

POSITIVE LOGITS

NetMessage

0.85

raq

0.79

iquette

0.77

inka

0.71

 Everest

0.71

 Banner

0.68

egu

0.68

Aff

0.67

ICAN

0.67

Pont

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.