INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ãĥį

-0.71

unin

-0.68

esm

-0.68

oaded

-0.67

LT

-0.67

 compe

-0.67

 disapp

-0.66

uner

-0.65

 provisional

-0.63

 recre

-0.62

POSITIVE LOGITS

1.07

1.04

yan

0.92

ya

0.91

yah

0.89

 Travels

0.87

yles

0.85

YA

0.82

legram

0.71

otaur

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.