INDEX

Explanations

The neuron strongly activates for the special start‐of‐sequence (“<bos>”) token.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tämän

0.40

 Colombeau

0.37

 ligação

0.36

 terkenal

0.36

 École

0.36

 holotype

0.36

 nahezu

0.35

 Campeonato

0.35

 nachhalt

0.34

 Lebih

0.34

POSITIVE LOGITS

if

0.38

while

0.36

package

0.34

set

0.34

///

0.34

 ৩৮

0.32

Log

0.32

On

0.32

```

0.31

Set

0.31

Activations Density 2.856%