INDEX

Explanations

modal verbs or "to" followed by verbs

This neuron responds strongly to first‐ and second‐person pronouns and related modals (“I,” “you,” “can,” “to,” “be,” etc.), effectively spotting personal statements or advice.

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

欤

-1.68

 eenvou

-1.66

spod

-1.48

 tejto

-1.47

🥽

-1.39

歟

-1.38

 aanv

-1.35

 oorspron

-1.35

 bijzondere

-1.34

 geweldige

-1.34

POSITIVE LOGITS

 just

1.68

 didn

1.38

それを

1.34

～」

1.28

虽然

1.23

try

1.23

その

1.22

แค่

1.20

？」

1.20

get

1.20

Activations Density 0.050%