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

ansky

-0.77

psons

-0.70

oday

-0.65

 vanish

-0.64

icter

-0.63

itability

-0.63

Woman

-0.63

Alien

-0.62

utton

-0.61

 Sahara

-0.61

POSITIVE LOGITS

kson

0.77

bent

0.76

high

0.73

framework

0.67

EStreamFrame

0.66

ãĤ©

0.65

tons

0.64

 Harmony

0.63

chall

0.62

wards

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.