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

ype

-0.71

 UCHIJ

-0.70

SPA

-0.70

BILITY

-0.68

eds

-0.67

UCH

-0.67

OLOGY

-0.66

VID

-0.65

 Archdemon

-0.64

ãĤ¼ãĤ¦ãĤ¹

-0.63

POSITIVE LOGITS

 pays

0.68

emetery

0.63

 climax

0.63

 proofs

0.62

ger

0.62

 Scouting

0.62

 Entreprene

0.60

cloud

0.60

 Colossus

0.59

hest

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.