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

istence

-0.78

umer

-0.75

arag

-0.72

nyder

-0.70

imens

-0.70

arij

-0.70

apters

-0.70

olars

-0.70

amar

-0.69

ingers

-0.68

POSITIVE LOGITS

 expectation

0.71

psi

0.66

PN

0.64

hood

0.64

 formation

0.63

EStream

0.63

 incumbent

0.63

ties

0.62

DOE

0.61

 motions

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.