INDEX

Explanations

literary analysis and technical errors

The neuron strongly activates on explicit mentions of literary‐critical and rhetorical terms—especially names of figures of speech and narrative devices like “metaphor,” “imagery,” “symbol(s),” “epic,” and “hero.”

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

 Anſ

-1.67

没能

-1.38

actéristi

-1.36

并未

-1.34

茕

-1.33

НИМА

-1.32

甞

-1.31

⛏

-1.30

𝔼

-1.30

 memutus

-1.30

POSITIVE LOGITS

in

1.84

not

1.61

 that

1.60

and

1.36

 usually

1.33

all

1.30

 each

1.29

1.22

 gave

1.20

 every

1.20

Activations Density 0.038%