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

ĸļ

-0.91

xual

-0.91

uddle

-0.76

ogens

-0.75

aryn

-0.75

idine

-0.74

ued

-0.71

ulton

-0.68

idays

-0.67

iko

-0.67

POSITIVE LOGITS

 Pillar

0.75

Phones

0.75

soDeliveryDate

0.63

 Grimm

0.61

 withdrawals

0.61

GE

0.60

psc

0.60

 generators

0.59

Flight

0.59

 Garrison

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.