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

VALUE

-0.85

¿

-0.78

Ī

-0.73

luaj

-0.71

ĭ

-0.70

 Malf

-0.69

ï¸

-0.67

eworthy

-0.66

CONCLUS

-0.65

 Fargo

-0.65

POSITIVE LOGITS

tein

0.69

renheit

0.69

wd

0.66

heit

0.65

isu

0.64

 Pillar

0.64

tin

0.63

baum

0.62

dan

0.60

gil

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.