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

 Perez

-0.80

 Chavez

-0.78

 Pruitt

-0.73

WTC

-0.72

 Font

-0.72

 Pose

-0.72

 Corpus

-0.70

 Pepe

-0.69

 Emirates

-0.66

 Sind

-0.65

POSITIVE LOGITS

etsy

0.87

ItemImage

0.86

romeda

0.80

iru

0.77

é¾įå

0.75

aspberry

0.70

atari

0.70

arb

0.69

hd

0.69

hai

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.