INDEX

Explanations

absence of something

The neuron activates on the word “without,” flagging negation by that specific term.

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

-2.31

 crucial

-1.82

The

-1.80

很想

-1.77

囁

-1.70

 maintenant

-1.63

何故

-1.62

 insane

-1.62

男朋友

-1.62

を見

-1.59

POSITIVE LOGITS

there

2.00

 ONLY

1.98

and

1.95

lossians

1.91

栻

1.82

事故

1.81

émoc

1.69

 This

1.68

 Some

1.66

1.65

Activations Density 0.036%