INDEX

Explanations

Hulk and Terminator references

The neuron fires on occurrences of prominent fictional character names (e.g. “Hulk,” “Bruce Banner,” “Terminator”) regardless of context.

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

 取り付け

-2.52

ولايات

-2.48

riezmann

-2.42

焄

-2.22

齪

-2.19

的那个

-2.09

jährige

-2.08

蕺

-2.06

 данного

-2.06

 laquelle

-2.06

POSITIVE LOGITS

𖣘

2.73

锒

2.66

寕

2.38

2.33

澔

2.33

棡

2.27

橚

2.22

雰

2.17

ными

2.17

 думал

2.16

Activations Density 0.001%