INDEX

Explanations

Python code examples

The neuron is primarily picking up on clause-level discourse pivots or contrastive conjunctions (e.g. “But”) that signal a sudden shift or counterpoint in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 차지

0.73

ạ

0.72

CTION

0.71

ರ

0.70

ं

0.69

 phir

0.68

Isopropyl

0.67

біль

0.66

Лі

0.66

 COLL

0.65

POSITIVE LOGITS

alchemy

0.68

 Rolle

0.66

 regex

0.64

 Andere

0.64

 feng

0.63

logos

0.62

 actores

0.62

otheby

0.62

版本的

0.61

 asign

0.60

Activations Density 0.000%