INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãģĻ

-0.78

ãĥ³ãĤ¸

-0.73

measures

-0.70

weight

-0.65

fort

-0.63

ãĤ¤ãĥĪ

-0.62

action

-0.61

iple

-0.61

aff

-0.61

ass

-0.61

POSITIVE LOGITS

utra

0.71

ingo

0.71

ixel

0.64

ocene

0.63

hon

0.62

 Polaris

0.62

esis

0.61

undo

0.60

Newsletter

0.60

lycer

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.