INDEX

Explanations

or otherwise + verb

The neuron is specialized in detecting the word “otherwise.”

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

 Both

-1.95

其他的

-1.71

果た

-1.68

どのような

-1.67

🎖

-1.61

似た

-1.59

&&

-1.59

\},

-1.58

 čase

-1.52

 slav

-1.52

POSITIVE LOGITS

2.55

</h3>

1.77

[]:

1.69

 called

1.64

1.63

1.59

est

1.59

ing

1.57

 feinen

1.57

стве

1.55

Activations Density 0.006%