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

 Awakening

-0.70

IDS

-0.69

éĥ

-0.66

ãĤ´

-0.66

ãĥĺãĥ©

-0.64

enture

-0.64

breaks

-0.63

ãĥİ

-0.62

bis

-0.61

 Series

-0.61

POSITIVE LOGITS

alogy

0.75

notations

0.70

iciary

0.69

ilies

0.66

erate

0.65

 Ripple

0.65

 signific

0.63

:{

0.61

oother

0.61

ability

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.