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

istically

-0.73

nick

-0.70

oji

-0.67

hattan

-0.67

 Advertisement

-0.62

iated

-0.62

istani

-0.61

nam

-0.61

ski

-0.61

 Sapphire

-0.60

POSITIVE LOGITS

Roll

0.67

obby

0.67

tnc

0.64

Front

0.64

 anim

0.63

 Lutheran

0.62

ugal

0.62

 Thom

0.61

ensor

0.61

uther

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.