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

cius

-0.70

 pron

-0.69

quickShipAvailable

-0.67

ALS

-0.66

 shines

-0.66

antry

-0.66

alist

-0.66

oting

-0.65

ISM

-0.65

orate

-0.64

POSITIVE LOGITS

¶æ

0.71

 Simulator

0.70

 Wast

0.70

 Proced

0.69

arthed

0.65

mid

0.65

letter

0.64

AFC

0.64

 Playoff

0.63

 Imag

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.