INDEX

Explanations

for-each loop iteration

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

 Engel

-0.11

 Lester

-0.10

elow

-0.10

_PID

-0.10

Lob

-0.08

lob

-0.08

 disarm

-0.08

PARSE

-0.08

æĲº

-0.08

 priv

-0.08

POSITIVE LOGITS

 foreach

0.25

foreach

0.24

 loop

0.19

_FOREACH

0.19

éģį

0.19

\tforeach

0.19

.forEach

0.19

.foreach

0.19

forEach

0.17

å¾ª

0.17

Activations Density 0.078%