esa-sceva/satcom-synth-qa-cot

---
dataset_name: satcom-synth-qa-cot
tags:
- satellite-communications
- synthetic-data
- reasoning
- chain-of-thought
- esa-sceva
language: en
license: apache-2.0
task_categories:
- question-answering
size_categories:
- 100K<n<1M
---

# esa-sceva/satcom-synth-qa-cot

## Summary
Synthetic chain-of-thought dataset with step-by-step reasoning QAs for satellite communications.

## Description
Generated with DeepSeek v3.2 and ChatGPT-5-Pro using topic-based prompts and structured answer templates.
Covers technical subjects such as link budget, Doppler shift, latency, antenna gain, and propagation delay.
Each answer includes reasoning steps, formulas, and explanations.

## Composition
- About 2.5k QAs
- JSONL format
- Fields: `question`, `answer`
- Language: English

## Intended use
Training or evaluating models for reasoning and problem solving in technical contexts.

## Quality control
- Deduplication and consistency filtering to ensure data integrity
- Token limit capped at 2048.
- Enforced step-by-step output structure

## Example
```python
from datasets import load_dataset
ds = load_dataset("esa-sceva/satcom-synth-qa-cot", split="train")
print(ds[0]["question"])
print(ds[0]["answer"])