INDEX

Explanations

function of

The neuron detects occurrences of “function” (and related dependency phrases like “as a function of” or “dependence”) that specify one variable’s relationship to another in scientific 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

なども

-1.41

などが

-1.20

 manteiga

-1.17

 помогают

-1.16

 combustível

-1.16

などは

-1.15

劵

-1.14

出场

-1.13

apeake

-1.13

 fijar

-1.13

POSITIVE LOGITS

of

2.58

for

1.34

konos

1.23

 increasing

1.23

тивного

1.16

Ведь

1.16

simov

1.15

Зем

1.14

at

1.14

 both

1.13

Activations Density 0.017%