INDEX

Explanations

your or yourself

This neuron detects imperative verbs that introduce commands or instructions (e.g., “Tell,” “Run,” “Use,” “Say,” “Make,” “Look,” “Try,” “Call”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

tentu

-1.40

PENUTUP

-1.38

 trebui

-1.38

olge

-1.32

 several

-1.32

ofed

-1.31

 dimas

-1.29

 akrab

-1.29

سوال

-1.25

って思

-1.24

POSITIVE LOGITS

 your

2.19

 yourself

1.63

 فهي

1.14

 például

1.13

你的

1.10

 podľa

1.09

 传统

1.09

把你

1.09

you

1.09

Activations Density 0.097%