INDEX

Explanations

code snippets and identifiers

The neuron strongly activates on numeric tokens—i.e. integer or numeric literals in code or text.

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

叭

-1.13

 biased

-1.05

AfterEach

-1.02

ゴ

-0.98

 engagé

-0.91

 deleteById

-0.91

ای

-0.91

 minde

-0.91

踪

-0.91

每一

-0.90

POSITIVE LOGITS

之际

1.05

いきなり

1.03

 what

1.01

ldon

1.00

meneu

0.98

 tungkol

0.96

相変わらず

0.94

 this

0.93

に行きました

0.90

んですか

0.90

Activations Density 0.004%