INDEX

Explanations

physics sciences study

The neuron detects definition‐style contexts, especially field or discipline names followed by a copula (“is the…”) in explanatory sentences.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 FanDuel

0.63

 overpowered

0.61

 menunggu

0.57

 ciné

0.56

 рабочий

0.55

 reinforced

0.55

 overpriced

0.54

 hyped

0.54

 když

0.54

 EGFR

0.54

POSITIVE LOGITS

 study

1.57

Study

1.48

study

1.46

 Study

1.43

 STUDY

1.38

 studying

1.29

 estudo

1.27

 estudia

1.27

 изучение

1.27

 estudio

1.24

Activations Density 0.139%