INDEX

Explanations

indicating consequences

The neuron is highly sensitive to discourse-linking words and phrases that introduce or connect clauses (e.g. “This,” “which,” “thus,” “finally,” “so”).

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

 estabiliz

-1.02

vasive

-0.98

 efeitos

-0.96

 publiée

-0.96

 transportar

-0.94

Tep

-0.92

 synes

-0.91

 procedimentos

-0.91

habitants

-0.90

 sagde

-0.89

POSITIVE LOGITS

 then

1.88

 thus

1.88

 resulting

1.77

 therefore

1.72

 thereby

1.71

 consequently

1.68

sequently

1.53

 ultimately

1.53

 result

1.49

 subsequently

1.47

Activations Density 0.075%