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

portation

-0.82

DragonMagazine

-0.78

ppa

-0.78

TABLE

-0.75

cards

-0.74

ĵĺ

-0.73

ERY

-0.72

wic

-0.71

utenberg

-0.70

CRIPTION

-0.70

POSITIVE LOGITS

 apex

0.74

 mouth

0.70

 medi

0.66

 abundance

0.62

enz

0.61

 choke

0.61

zeb

0.61

 hostage

0.59

 clicks

0.59

 head

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.