SharkSpicy/wikidata

SR-KI Dataset

The dataset accompanying SR-KI: Scalable and Real-Time Knowledge Integration into LLMs via Supervised Attention (AAAI 2026).

Overview

The SR-KI Dataset provides Chinese question-answering data for training and evaluating the supervised-attention knowledge-integration method introduced in the SR-KI paper. Each example pairs a question and answer with the supporting knowledge and its corresponding material identifier, enabling models to produce grounded answers with explicit source attribution.

For details about the method, experimental setup, and implementation, please refer to the paper and the official GitHub repository.

Dataset Structure

The dataset contains train and test splits. Each record includes the following fields:

Field	Description
doc_id	Identifier(s) of the supporting material.
nodes	Knowledge elements associated with the example.
Q	Original question.
A	Reference answer.
kb_info	Structured supporting knowledge and material metadata.
qa_type	Question-answering category.
Q_format	Formatted model input or instruction.
A_format	Formatted target answer with source attribution.

Citation

If you find this dataset helpful in your research, we would kindly appreciate a citation:

@misc{yu2025srkiscalablerealtimeknowledge,
  title         = {SR-KI: Scalable and Real-Time Knowledge Integration into LLMs via Supervised Attention},
  author        = {Bohan Yu and Wei Huang and Kang Liu},
  year          = {2025},
  eprint        = {2511.06446},
  archivePrefix = {arXiv},
  primaryClass  = {cs.CL},
  url           = {https://arxiv.org/abs/2511.06446}
}