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

etsk

-0.83

uers

-0.73

xon

-0.72

regor

-0.72

bley

-0.71

uben

-0.70

 Lans

-0.70

assic

-0.68

 convenience

-0.68

vir

-0.67

POSITIVE LOGITS

..................

0.65

.''.

0.63

lining

0.61

rock

0.61

 liability

0.59

punk

0.59

Pinterest

0.58

 Illegal

0.58

?ãĢį

0.57

 OPEC

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.