INDEX

Explanations

Analyze others reading

The neuron strongly activates on verbs and phrases that specify student learning outcomes or instructional objectives (e.g. identify, explain, analyze, understand) in educational or lesson‐plan 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

 really

-1.02

 onwards

-0.96

 roughly

-0.93

 absolutely

-0.93

 completely

-0.92

 terribly

-0.92

 exactly

-0.91

 truly

-0.90

 whopping

-0.89

 actually

-0.85

POSITIVE LOGITS

 bytu

1.00

 darse

0.97

Jp

0.96

zmán

0.90

emate

0.89

 distinctive

0.88

nières

0.87

NavLink

0.86

cycl

0.85

 Fauci

0.85

Activations Density 0.015%