INDEX

Explanations

responsibility and blame

This neuron responds to statements where someone explicitly accepts responsibility or “takes accountability.”

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

 softening

0.57

 উপলক্ষে

0.56

|}\

0.54

 চলিয়া

0.53

presets

0.52

 tige

0.52

 बाथरूम

0.51

ហារ

0.51

 antigu

0.50

 multicolor

0.49

POSITIVE LOGITS

 blame

3.20

 responsibility

3.05

responsibility

2.92

 Responsibility

2.81

 blames

2.77

 blamed

2.72

責任

2.72

 responsabilité

2.67

 responsabilidad

2.66

 blaming

2.66

Activations Density 0.469%