INDEX

Explanations

official actions and status

The neuron strongly activates on words naming formal, bureaucratic or administrative actions and processes (e.g. applied, submitted, approved, nomination, notice).

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

否认

-1.19

と思われます

-1.14

もありますが

-1.09

bouncycastle

-1.08

でしょうね

-1.05

 solamente

-1.05

Its

-1.05

これ

-1.04

 afore

-1.03

 Nineteen

-1.03

POSITIVE LOGITS

for

1.59

by

1.51

on

1.48

 your

1.22

the

1.15

屋さん

1.13

ULATIONS

1.12

latimes

1.09

AdapterView

1.09

FOR

1.09

Activations Density 0.079%