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

leaf

-0.87

tes

-0.75

catentry

-0.73

oval

-0.72

zens

-0.72

Accessory

-0.71

GV

-0.70

document

-0.70

IPP

-0.68

 lett

-0.68

POSITIVE LOGITS

acebook

0.71

 Prelude

0.69

 Rail

0.67

Ü

0.66

eways

0.64

 Revel

0.63

lain

0.63

 Clip

0.63

 reception

0.62

hower

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.