INDEX

Explanations

clause separators and conjunctions

The neuron strongly responds to mentions of volatile vs. non-volatile memory types and related memory‐category terms (e.g. RAM, SRAM, DRAM).

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

-1.71

and

-1.59

 which

-1.38

 стан

-1.35

on

-1.25

 regarding

-1.23

ğinde

-1.22

 myös

-1.20

猡

-1.16

 various

-1.15

POSITIVE LOGITS

ఝ

1.44

 religieuses

1.44

橚

1.43

 вашего

1.41

 geprü

1.41

 angenommen

1.39

 zufried

1.39

 einzelnen

1.38

 anzunehmen

1.37

 hilfreich

1.36

Activations Density 0.536%