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

OND

-0.77

ãĥ¼ãĥĨãĤ£

-0.73

 suspic

-0.73

anwhile

-0.72

unts

-0.70

mination

-0.70

MpServer

-0.68

oop

-0.67

glers

-0.67

achu

-0.66

POSITIVE LOGITS

 Atom

0.68

 Silver

0.67

her

0.66

fty

0.66

ami

0.65

thal

0.65

indu

0.63

given

0.63

der

0.63

 Pall

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.