INDEX

Explanations

input or example markers

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

oga

-0.09

alue

-0.09

nameof

-0.09

 Aggregate

-0.08

bery

-0.08

ia

-0.08

 HOLDER

-0.08

Ð¸ÑĤÐµ

-0.08

 Schw

-0.08

leyin

-0.08

POSITIVE LOGITS

 input

0.18

 example

0.17

 Input

0.16

 Example

0.15

Input

0.14

Example

0.13

è¾ĵåħ¥

0.13

input

0.13

-input

0.13

ä¾ĭ

0.12

Activations Density 0.140%