INDEX

Explanations

testing assertions and equality checks

This neuron detects unit‐test assertion and expectation keywords (e.g. contains, toBe/toEqual, assert, eq, deepEqual, CHECK_EQUAL, etc.).

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

 seemed

-1.16



-1.10

 makes

-1.09

ஶ

-1.04

 seems

-1.04

 ilha

-1.03

 whose

-0.98

。《

-0.98

as

-0.96

 seemingly

-0.96

POSITIVE LOGITS

 expected

1.52

expect

1.35

 expect

1.34

expected

1.32

 Expected

1.24

 EXPECT

1.23

 expectation

1.20

 همان

1.19

 ожида

1.15

 esperado

1.13

Activations Density 0.003%