INDEX

Explanations

has been

The neuron is tuned to pick out auxiliary “have/has” (and related perfect‐aspect auxiliaries like “been”) in sentences.

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

and

-1.10

matics

-1.00

tiness

-0.98

umą

-0.98

🤨

-0.95

😕

-0.95

inį

-0.94

adone

-0.91

BIE

-0.91

 ihan

-0.90

POSITIVE LOGITS

 been

3.14

 shown

1.96

 showed

1.80

 become

1.79

 demonstrated

1.70

 recently

1.60

 proved

1.53

 also

1.42

1.37

 menunjukkan

1.34

Activations Density 0.140%