INDEX

Explanations

is or isValid checks

This neuron detects occurrences of the token “is” (especially in Boolean‐predicate or validity‐check identifiers) in source code.

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

 constater

-1.23

 quelques

-1.21

 Meksiko

-1.20

 بدون

-1.20

 maksi

-1.20

 corriger

-1.19

 siè

-1.16

きれいな

-1.16

incredible

-1.16

tische

-1.13

POSITIVE LOGITS

any

1.48

all

1.38

 this

1.27

 still

1.25

as

1.25

has

1.24

 there

1.23

 having

1.23

had

1.22

 under

1.22

Activations Density 0.018%