INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

cells

-0.71

egg

-0.69

dt

-0.69

range

-0.67

tes

-0.66

te

-0.65

 Tray

-0.65

ele

-0.63

orus

-0.63

 Eternity

-0.63

POSITIVE LOGITS

itivity

0.88

assad

0.73

xual

0.71

 shedding

0.71

 Flavoring

0.70

 causation

0.66

 attRot

0.64

 cardio

0.64

lations

0.63

henko

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.