INDEX

Explanations

subatomic particles and electrons

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

ãĥĭãĤ¢

-0.09

suz

-0.09

 atmos

-0.09

IGO

-0.09

ÐºÑĥÐ»ÑĮ

-0.09

isan

-0.09

éľŀ

-0.09

 Thunder

-0.09

 Torch

-0.09

 dynam

-0.09

POSITIVE LOGITS

 particles

0.23

 electrons

0.21

 electron

0.19

 neut

0.18

 particle

0.18

 neutr

0.16

ç²Ĵ

0.15

particles

0.15

éĽ»åŃĲ

0.15

qu

0.14

Activations Density 0.060%