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

opter

-0.76

pair

-0.76

hack

-0.75

Cass

-0.73

ault

-0.68

uch

-0.68

pas

-0.67

Attach

-0.67

lik

-0.67

pter

-0.67

POSITIVE LOGITS

 Skydragon

0.79

agnar

0.79

agna

0.69

 swast

0.67

 Scientology

0.67

 ethnicity

0.67

 athlet

0.66

 Barkley

0.64

 incarcer

0.63

 Scandinav

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.