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

ãĥ´ãĤ¡

-0.71

 recall

-0.68

repair

-0.65

advertisement

-0.65

umen

-0.63

anium

-0.63

 packages

-0.62

conserv

-0.62

parts

-0.60

archives

-0.60

POSITIVE LOGITS

oulos

0.89

iets

0.78

hovah

0.73

 Barkley

0.72

 Dawkins

0.66

 surname

0.64

hedon

0.64

 creed

0.64

SEE

0.64

HBO

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.