INDEX

Explanations

studies modifying region interventions

The neuron strongly activates on longer, multi-syllabic content words (e.g. “skeletons,” “spectroscopic,” “monoliths,” “learning,” etc.).

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

but

-1.70

 نداشت

-1.34

빡

-1.33

𓁹

-1.27

﹍﹍

-1.26

と話

-1.25

 gamle

-1.23

decided

-1.23

 aproxim

-1.21

 obat

-1.20

POSITIVE LOGITS

on

1.48

 additional

1.30

new

1.27

ba

1.23

of

1.22

co

1.19

 escola

1.18

an

1.17

1.16

 rész

1.16

Activations Density 0.160%