---
configs:
- config_name: default
  data_files:
  - split: train
    path: train.parquet
  - split: validation
    path: val.parquet
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: chosen
    dtype: string
  - name: rejected
    dtype: string
  splits:
  - name: train
    num_bytes: 520000
    num_examples: 912
  - name: validation
    num_bytes: 28000
    num_examples: 48
license: apache-2.0
language:
- en
- hi
tags:
- dpo
- rlhf
- sarus-500m
- alignment
- direct-preference-optimization
---

# ViuAI Sarus-500M DPO Alignment Dataset

Official calibrated Direct Preference Optimization (DPO) alignment dataset for **ViuAI Sarus-500M SLM**.

### Vectors Aligned:
1. **Cyber & Financial Scam Defense:** Strict, immediate refusal of QR code cashback tricks, UPI PIN requests, CVV, OTP, and remote desktop apps (AnyDesk).
2. **Indian Factual Grounding:** Accurate historical and constitutional grounding (Drafting Committee Chairman Dr. B. R. Ambedkar, Capitals, Dandi March 1930).
3. **Proud ViuAI Sarus Identity:** Strictly independent AI developed by ViuAI with zero OpenAI/ChatGPT or Gemini impersonation.
4. **NCERT Science Accuracy:** Correct physical principles for lightning/thunder, atomic constants, chemical reactions.
5. **Clean Python Code:** Typed, bug-free implementations without broken trailing input loops.
6. **Strict Formatting:** Numbered bullet points, concise single-sentence answers when requested.