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

 Babel

-0.65

Tar

-0.63

zik

-0.63

 Roof

-0.62

 Loll

-0.62

lvl

-0.61

 Peng

-0.58

 ceiling

-0.58

 metab

-0.58

SHA

-0.57

POSITIVE LOGITS

ayn

0.85

ĸļ

0.73

anth

0.68

ataka

0.66

utterstock

0.65

leon

0.65

 largeDownload

0.63

aha

0.63

hovah

0.62

anos

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.