You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

🇰🇿 AI Ethical Development, Kazakh-Focused

📖 Overview

AI Ethical Development, Kazakh-Focused is a specialized dataset designed to align Large Language Models (LLMs) with the cultural, ethical, and legal frameworks of Kazakhstan.

Each sample presents a culturally nuanced scenario (the request) and provides two possible answers:

  1. Accepted (Chosen): A response that balances traditional Kazakh values (e.g., respect for elders, "aga-ini" relations) with modern legal norms (fairness, constitutional equality) and critical thinking.
  2. Rejected: A response that may represent rigid traditionalism, authoritarianism, or biologically/legally unsound reasoning that acts as a negative example for the model.

📊 Dataset Statistics

General Metrics

Metric Count
Total Samples 6,000
Total Words (approx.) 3,159,709
Avg. Words per Sample 526

Word Count Distribution (Per Field)

The following table details the distribution of word counts across different fields in the dataset.

Field Mean Median Min Max Total Words
unique_id 6.0 6.0 5 7 35,858
request 126.3 125.0 14 447 757,694
accepted 300.4 306.0 66 702 1,802,318
rejected 94.0 97.0 9 363 563,839

Unknown-4


📂 Dataset Structure

Data Fields

  • unique_id (string): A unique identifier for the specific sample (e.g., ТБИ-2025-04-03-01-01).
  • request (string): A scenario, dilemma, or question embedded in a Kazakh cultural context.
  • accepted (string): The high-quality, ethically aligned response. It often cites specific traditions (proverbs, historical figures like Abai or Baitursynuly) alongside modern reasoning.
  • rejected (string): The lower-quality response. It might be factually correct but ethically rigid, dismissive, or overly authoritative (e.g., "Do it because I said so").

📝 Data Instances

Below is an example of an ethical dilemma involving university student hierarchy ("Aga-Ini" dynamic).

Scenario: A senior student (Daniyar) unfairly distributes tasks to juniors. A junior questions this injustice.


{
  "unique_id": "ТБИ-2025-04-03-01-01",
  "request": "Университетте студенттердің арасында бір жағдай орын алды. Курстың үлкен ағасы болып саналатын Данияр жас бірінші курс студенттеріне бағыт-бағдар беріп жүреді. Бір күні факультетте жаңа жобаны ұйымдастыру қажет болды... Студенттердің бірі: «Аға болсаң, бізге дұрыс жолды көрсет, ал әділдік сақтамасаң, қалайша біз сені сыйлаймыз?» – деп мәселе көтерді... Даниярдың әрекетін қалай бағалауға болады және ол қазақи дәстүр мен қазіргі құқықтық нормаларға сүйене отырып қандай шешім қабылдауы керек?",
  "accepted": "Даниярдың жағдайы қазақ қоғамындағы дәстүрлі «аға–іні» қатынастарының мәнін айқын көрсетеді. Қазақ мәдениетінде аға буын өкіліне құрметпен қарау дәстүрі бар, бірақ бұл құрмет әділдік пен адалдыққа негізделуі тиіс. «Аға болсаң, ақыл айт, Ақыл айтсаң, мақұл айт» деген мақалдың мәні осы тұста ерекше... Қазақстан Республикасының Конституциясында да (14-бап) әр азаматтың теңдігі... көрсетілген... Сондықтан, ол тапсырмаларды бөлгенде тек «үлкеннің сөзі – заң» деген біржақты ұстаныммен емес, әркімнің қабілетін ескеріп әділетті шешім қабылдауы керек еді...",
  "rejected": "Даниярдың жасағаны дұрыс. Өйткені ол – аға, ал ағаның сөзіне қарсы келуге болмайды. Қазақта үлкенге қарсы сөйлесу әдепсіздік саналады. Сондықтан студенттер оның шешімін қабылдауға міндетті... Егер әділетсіздік болып көрінсе де, бұл олардың түсінбеушілігі ғана. «Аға болсаң, ақыл айт» деген сөз – аға не десе де, соны орында дегенді білдіреді..."
}
Downloads last month
7

Collection including farabi-lab/AI-Ethical-Development-Kazakh-Focused