INDEX

Explanations

tell me

The neuron fires strongly on verbs of speech or reporting—words like “told,” “taught,” “recounted,” “explained,” and “tell” that indicate someone speaking or narrating.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

<0x80>

1.30

1.26

1.23

1.19

्त

1.16

்க்கு

1.12

1.09

 usur

1.05

ぅ

1.05

ない

1.05

POSITIVE LOGITS

 وضعیت

1.15

实话

1.12

 TELL

1.09

 Ива

1.07

ynthesis

1.02

 सांगित

1.02

 kisah

1.02

 सांगितलं

1.02

 கூறியதாவது

1.02

 mitä

1.01

Activations Density 0.129%