INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ajor

-0.78

 Warsaw

-0.77

fleet

-0.74

plex

-0.69

 squared

-0.68

 Wesley

-0.66

awar

-0.65

 Enterprise

-0.65

negie

-0.64

conom

-0.64

POSITIVE LOGITS

 ethics

0.73

 Investig

0.67

arna

0.66

 contrary

0.66

ITCH

0.66

 Inquiry

0.65

rahim

0.64

 perjury

0.63

ALSE

0.63

 FOIA

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.