INDEX

Explanations

let's

The neuron fires strongly on occurrences of “let” (including “let’s” and “let me”), i.e. phrases introducing a directive or invitation.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ির

1.70

 положения

1.40

 supremum

1.31

ﺐ

1.26

িতে

1.24

 nace

1.23

。

1.23

べて

1.20

‌ای

1.18

ﻂ

1.17

POSITIVE LOGITS

tted

1.62

ts

1.55

tt

1.50

tter

1.50

ש

1.50

ν

1.48

tting

1.48

س

1.46

ttet

1.44

 intérieure

1.38

Activations Density 0.042%