INDEX

Explanations

subject plus verb or contrast

The neuron strongly activates on common personal pronouns (I, it, they, all) and frequent auxiliary/modal verbs (is, would, does, just), i.e. high-frequency function words.

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

 ikke

-1.26

でない

-1.18

 nicht

-1.14

 אין

-1.14

不再

-1.10

毫不

-1.10

 không

-1.09

 keine

-1.05

格外

-1.00

 artık

-0.99

POSITIVE LOGITS

 instead

3.52

 vielmehr

3.08

 simply

2.88

Instead

2.75

 Instead

2.73

 merely

2.66

而是

2.66

むしろ

2.55

nor

2.34

 invece

2.28

Activations Density 0.058%