INDEX

Explanations

admission, exclusion, evidence

The neuron specializes in spotting legal rule or statute references—i.e. the numbered sections, rules, and evidence‐act citations in legal text.

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

Sv

-0.77

 unoccupied

-0.75

Animator

-0.69

 Foot

-0.68

𝓸

-0.68

DSL

-0.67

发

-0.67

 Delegation

-0.67

oğ

-0.66

 sinon

-0.66

POSITIVE LOGITS

 admis

2.25

 admission

1.80

 exclusion

1.68

 admissible

1.63

 exclude

1.58

 evidence

1.57

 Admission

1.50

exclude

1.48

Admission

1.48

exclusion

1.46

Activations Density 0.042%