INDEX

Explanations

explaining and instructing

The neuron detects verbs that denote giving instructions or advice (e.g., explain, tell, instruct, ask, suggest).

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

on

-1.30

 only

-1.13

 from

-1.12

 depuis

-1.05

-1.01

all

-1.00

 keja

-0.97

↵↵

-0.96

versiti

-0.95

ἑ

-0.95

POSITIVE LOGITS

犼

1.22

خفض

1.20

ніка

1.20

 svůj

1.20

nước

1.20

身邊

1.20

brtc

1.20

тельности

1.19

calça

1.18

 samenwerking

1.16

Activations Density 0.049%