INDEX

Explanations

changing mind or position

The neuron detects words and phrases signaling a change or reversal of stance (e.g. change, changed, reversal).

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

 foremost

-1.48

наче

-1.45

 sogen

-1.37

与此同时

-1.34

 besonderes

-1.30

 beleza

-1.29

鹛

-1.28

ꜥ

-1.27

 biraz

-1.27

 serene

-1.26

POSITIVE LOGITS

]='

1.47

 igenom

1.37

，“

1.36

 به

1.34

żenie

1.34

You

1.29

use

1.28

 undergoes

1.26

。』

1.26

乀

1.25

Activations Density 0.014%