INDEX

Explanations

pronouns in relation to others

The neuron responds strongly to semantically rich content words (nouns, verbs, and adjectives carrying factual or narrative detail) rather than to common function words.

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

它们的

0.96

她们

0.95

พวก

0.89

=$_

0.84

 રહ્યું

0.82

டுகின்றன

0.82

 acompañada

0.78

 ими

0.75

તું

0.71

illée

0.71

POSITIVE LOGITS

 그는

4.81

他在

4.50

he

4.47

 він

4.47

him

4.41

他也

4.38

เขา

4.28

 เขา

4.28

彼は

4.22

 তিনি

4.19

Activations Density 3.797%