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

iday

-0.81

ãĤ¨ãĥ«

-0.80

amines

-0.79

iversary

-0.76

ival

-0.76

aceous

-0.75

insk

-0.74

arios

-0.71

lins

-0.70

apes

-0.70

POSITIVE LOGITS

 acceptable

0.69

 brake

0.63

 pret

0.62

 neglig

0.59

 gull

0.59

TOD

0.59

 toler

0.58

 nominal

0.58

AFTA

0.58

Eh

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.