INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãģ®æ

-0.77

 podcast

-0.76

GROUP

-0.74

DragonMagazine

-0.71

ourcing

-0.70

 sshd

-0.66

KEN

-0.65

ursday

-0.64

pez

-0.64

 corrid

-0.63

POSITIVE LOGITS

Tel

0.76

ËĪ

0.68

hap

0.67

 Shap

0.67

cap

0.66

mort

0.65

clock

0.65

 grips

0.65

birth

0.63

Atk

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.