INDEX

Explanations

module object layer

The neuron fires on occurrences of the substring “nn” in code, e.g. references to PyTorch’s torch.nn module.

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

鯛

-0.88

それでも

-0.86

 nehme

-0.81

槎

-0.81

 computing

-0.79

 however

-0.79

が上が

-0.79

켭

-0.78

减少

-0.77

idx

-0.77

POSITIVE LOGITS

 Grit

0.90

穴

0.88

0.87

ध्य

0.84

esség

0.84

Fluorescence

0.83



0.82

既然

0.82

Sequential

0.82

Click

0.82

Activations Density 0.003%