INDEX

Explanations

groundbreaking studies and events

The neuron detects the adjective “groundbreaking” (even when split across tokens).

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

</i>

-2.33

 menangkap

-2.20

'>{

-2.19

-1.94

ra

-1.90

两

-1.87

</th>

-1.87

ться

-1.84

wwww

-1.81

";

-1.80

POSITIVE LOGITS

</h4>

2.28

</em>

2.28

 they

2.11

2.03

and

1.88

 alongside

1.88

・・。

1.85

<em>

1.83

无声

1.77

沨

1.76

Activations Density 0.005%