INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Liter

-0.69

 targ

-0.66

ibal

-0.60

 Discussion

-0.59

endix

-0.59

dyn

-0.56

paio

-0.56

 remembrance

-0.56

 predec

-0.55

 emphasis

-0.55

POSITIVE LOGITS

anan

0.72

window

0.71

lihood

0.70

 gallery

0.69

utz

0.69

 srfAttach

0.68

ãĤĬ

0.68

PRO

0.67

antes

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.