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

 Colleges

-0.73

manager

-0.70

Hill

-0.67

equal

-0.61

 Citation

-0.60

 lists

-0.59

HR

-0.58

Work

-0.57

Williams

-0.57

mill

-0.56

POSITIVE LOGITS

incinn

0.84

inarily

0.81

dinand

0.78

igham

0.75

initely

0.73

hower

0.71

enes

0.71

itic

0.71

ivably

0.71

OTO

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.