INDEX

Explanations

output generation or destination

The neuron is triggered by tokens naming file‐ or data‐I/O operations (e.g. open, write, output, to) signaling reading from or writing to files.

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

 from

-1.04

 additional

-1.04

 only

-0.93

できません

-0.92

 with

-0.91

 earlier

-0.90

 additionally

-0.88

 onResume

-0.86

 frustrating

-0.86

 That

-0.85

POSITIVE LOGITS

 OUTPUT

1.24

 output

1.20

OUTPUT

1.18

output

1.16

 outputs

1.05

出力

1.01

 Output

0.97

Output

0.96

输出

0.94

OUT

0.93

Activations Density 0.050%