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

Bra

-0.78

 Xuan

-0.78

managed

-0.65

ä¹ĭ

-0.64

Champ

-0.64

 Yong

-0.63

 Â¥

-0.63

 treated

-0.62

Bat

-0.62

èª

-0.62

POSITIVE LOGITS

aminer

0.92

lycer

0.80

sylvania

0.75

ilege

0.70

lando

0.69

otine

0.68

esty

0.68

 outweigh

0.68

milo

0.67

enhagen

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.