INDEX

Explanations

summarizing statements and data

This neuron responds to cues indicating that the text is summarizing or paraphrasing information (e.g. words like “what,” “summary,” or “to paraphrase”).

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

ids

-1.30

 seperti

-1.22

 quando

-1.19

ONLY

-1.18

such

-1.15

Seznam

-1.13

["

-1.10

 ширина

-1.10

sl

-1.09

zembro

-1.09

POSITIVE LOGITS

 that

1.32

 both

1.26

 enfance

1.23

 then

1.15

 damaligen

1.11

ってこと

1.10

how

1.09

 yaka

1.09

TRIBUTES

1.06

 damal

1.06

Activations Density 0.097%