INDEX

Explanations

The neuron is picking up on self-referential or opinion-marking language (e.g. “I,” “maybe,” “because,” “someone”) that signals subjective commentary.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 apparently

1.29

 evidently

1.26

 praticamente

1.21

 usually

1.20

 basically

1.13

 presumably

1.12

 obviously

1.11

 undoubtedly

1.10

幾乎

1.07

 clearly

1.06

POSITIVE LOGITS

 algún

1.23

 someday

1.16

 কিছুটা

1.10

もう少し

1.08

 κάπο

1.07

 nějak

1.05

 einigen

1.04

 qualche

1.02

 इसीलिए

1.01

ielleicht

1.01

Activations Density 1.092%