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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gary

-0.82

cart

-0.78

hold

-0.78

nexus

-0.72

holder

-0.70

nell

-0.70

Coun

-0.69

ously

-0.68

ually

-0.65

cape

-0.65

POSITIVE LOGITS

 Bard

0.69

ĺħ

0.65

 Shib

0.63

Ã¡n

0.62

Rey

0.61

Riy

0.61

 Guan

0.60

 Hust

0.58

 Rocket

0.58

 Somers

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.