INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Abstract

-0.80

oute

-0.75

udeau

-0.75

umbai

-0.72

cro

-0.69

mean

-0.69

manac

-0.69

addafi

-0.68

bh

-0.67

phasis

-0.65

POSITIVE LOGITS

 misunder

0.74

Īè

0.69

DER

0.69

VOL

0.66

 Corpor

0.65

ANK

0.63

zek

0.63

encies

0.63

OWN

0.62

amar

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.