INDEX

Explanations

acid, sour, tart

The neuron fires on words describing a sour or tart taste (e.g., “tart,” “tangy,” “acid,” “sour”).

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

who

-1.36

Karten

-1.34

älde

-1.33

結局

-1.31

 насељу

-1.27

ORDIN

-1.26

といった

-1.21

戥

-1.21

awm

-1.21

ellants

-1.20

POSITIVE LOGITS

ing

1.63

 antichi

1.45

 både

1.38

1.35

 acid

1.34

 advertencia

1.33

 confirmación

1.30

 داشتند

1.28

дні

1.27

 indique

1.25

Activations Density 0.107%