INDEX

Explanations

dict.items() calls

The neuron fires on loop‐iteration identifiers and mapping access calls (e.g. “.items()”, “key”, “keys”, “k”)—i.e. elements involved in iterating over dictionaries or arrays.

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

in

-1.27

 when

-1.10

 after

-1.08

one

-1.06

 with

-1.04

at

-1.04

 from

-1.03

for

-1.02

two

-0.94

as

-0.92

POSITIVE LOGITS

ビックリ

1.18

びっくり

1.09

 gäng

1.03

 développer

1.02

どれだけ

0.98

czu

0.98

だら

0.98

 réserver

0.97

うら

0.96

 Rücksicht

0.94

Activations Density 0.003%