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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 teasp

-0.83

 satell

-0.75

ombie

-0.68

claimed

-0.68

capital

-0.67

consumer

-0.67

phy

-0.67

mop

-0.66

entric

-0.66

mercial

-0.66

POSITIVE LOGITS

ÂłÂłÂł

0.76

 Isles

0.73

++++++++++++++++

0.69

 Blumenthal

0.69

een

0.68

 Aram

0.66

 Serge

0.63

--+

0.62

rar

0.61

 Uruguay

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.