INDEX

Explanations

code block starters

The neuron activates on numeric code tokens—particularly standalone numbers and related assignment symbols.

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

quartered

0.61

 পরিবর্তিত

0.59

Branches

0.59

 هیڅ

0.59

ರುವುದರಿಂದ

0.59

াউন

0.58

 ADMINISTRATIVE

0.57

LLI

0.57

 Branches

0.57

 চলিয়া

0.57

POSITIVE LOGITS

示例

0.77

<body>

0.77

 bạn

0.71

 kullanıcı

0.71

 của

0.69

<html>

0.68

...');

0.67

{\

0.66

 leta

0.66

let

0.65

Activations Density 0.011%