INDEX

Explanations

like a

The neuron fires on colloquial simile or comparison cues—especially the discourse marker “like” (and allied phrases such as “sounds like” or “seems like”).

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

니다

1.08

습니다

1.05

드

1.00

Fragments

0.98

دى

0.93

ն

0.92

 Indexes

0.92

 дроб

0.92

ν

0.92

ง

0.91

POSITIVE LOGITS

it

1.57

 څنګه

1.41

 geschaffen

1.24

 folgt

1.23

 savent

1.17

س

1.17

ти

1.16

िक

1.14

fantastic

1.13

ould

1.11

Activations Density 0.033%