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

pend

-0.69

milo

-0.63

 Newton

-0.62

Downloadha

-0.62

formed

-0.62

thinkable

-0.61

sed

-0.60

months

-0.60

pez

-0.60

 paed

-0.59

POSITIVE LOGITS

enegger

0.85

kie

0.70

 guiActive

0.70

uren

0.69

ADVERTISEMENT

0.68

eous

0.66

aez

0.66

Ŀ

0.64

nen

0.63

 indeed

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.