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

ability

-0.75

li

-0.73

defined

-0.72

inational

-0.71

ichick

-0.71

purpose

-0.67

orate

-0.65

Pwr

-0.65

quartered

-0.64

pg

-0.64

POSITIVE LOGITS

 Freak

0.72

Studio

0.70

MFT

0.70

asca

0.70

owa

0.66

 Auditor

0.66

 Actress

0.65

 Concert

0.65

 before

0.65

ublic

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.