INDEX

Explanations

reading file content

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

ysa

-0.09

anik

-0.09

 sway

-0.09

ilor

-0.09

 hydr

-0.09

 Alerts

-0.08

oler

-0.08

 blink

-0.08

 Falcon

-0.08

hydr

-0.08

POSITIVE LOGITS

.read

0.21

.ReadToEnd

0.20

.readlines

0.19

 read

0.14

.ReadAll

0.13

 readline

0.13

.readline

0.13

.write

0.12

.close

0.11

.ReadAllText

0.11

Activations Density 0.030%