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

rational

-0.78

jury

-0.75

 boycott

-0.72

wcs

-0.68

sonian

-0.68

 boycot

-0.68

oult

-0.67

liga

-0.67

displayText

-0.66

amaz

-0.66

POSITIVE LOGITS

aps

0.75

 talents

0.73

fits

0.72

 Brist

0.72

 Noon

0.70

 inher

0.67

ipers

0.66

 Infinity

0.65

 babys

0.63

ights

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.