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

MpServer

-0.77

 agric

-0.73

wagen

-0.73

oufl

-0.71

Cth

-0.68

erers

-0.68

ieu

-0.67

bred

-0.67

RPG

-0.66

jar

-0.65

POSITIVE LOGITS

enza

0.70

 Theft

0.69

ivism

0.64

protect

0.63

 Vale

0.63

 Streets

0.63

leans

0.63

 Intercept

0.62

 Views

0.59

 Station

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.