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

chance

-0.81

 horizont

-0.76

Osw

-0.72

endor

-0.71

stru

-0.70

Kn

-0.68

zb

-0.68

abs

-0.67

Unix

-0.66

nr

-0.66

POSITIVE LOGITS

quickShipAvailable

0.74

 Editors

0.64

 accent

0.61

llor

0.60

reon

0.60

 motto

0.59

 modifier

0.58

 Dhabi

0.57

 dare

0.57

 adjective

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.