INDEX

Explanations

introduction related phrases

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ToEnd

-0.10

igest

-0.09

UpDown

-0.08

asant

-0.08

 fluores

-0.08

 gÃ¶n

-0.08

/includes

-0.08

_inp

-0.08

ç§Ģ

-0.08

æĽ

-0.08

POSITIVE LOGITS

 introduction

0.45

 Introduction

0.39

Introduction

0.36

 introdu

0.35

 intro

0.35

 introduce

0.28

intro

0.26

 opening

0.26

 introducing

0.25

ä»ĭç»į

0.24

Activations Density 0.245%