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

ocrat

-0.68

ollow

-0.67

Aly

-0.66

SOC

-0.66

oslov

-0.66

 Mellon

-0.65

ette

-0.64

apult

-0.62

 Jinping

-0.62

hern

-0.62

POSITIVE LOGITS

resist

0.67

weet

0.65

flies

0.65

ntil

0.64

mys

0.63

 Monetary

0.63

asio

0.63

ãĥĺãĥ©

0.62

otto

0.62

[|

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.