GPE

This repository is the official code implementation of:

GPE: Evaluating Robust Evidence Aggregation for Fact Verification under Controllable GEO-Style Poisoning

Experimental project. The code and benchmark are provided for research use and may contain bugs. Please consult the paper for the full methodology, experimental settings, and interpretation of results.

Install

python -m pip install gpe-eval

For development, clone the repository and create the included Conda environment:

conda env create -f environment.yml
conda activate gpe

Use

The graph and retrieval example uses only the bundled data:

python exp/run_graph_and_retrieval.py \
  --entity BBC \
  --claim-id benchmark-000001 \
  --query "steel production emissions" \
  --top-k 3

To run the claim-local or global retrieval evaluation smoke test, configure LLM_API_KEY in .env, then run one of:

python exp/run_claim_retrieval_evaluation.py \
  --method direct_evidence \
  --smoke-test \
  --threads 1

python exp/run_global_retrieval_evaluation.py \
  --method direct_evidence \
  --smoke-test \
  --threads 1

Citation

If you find this repository or the GPE benchmark useful, please cite:

@article{wang2026gpe,
  title={GPE: Evaluating Robust Evidence Aggregation for Fact Verification under Controllable GEO-Style Poisoning},
  author={Wang, Zhaoqi and Zhang, Zijian and Yuan, Xiaomei and Kou, Pengtao and Liu, Jiamou and Li, Zhen and Zhu, Liehuang},
  journal={arXiv preprint arXiv:2607.20730},
  year={2026}
}

License

This project is licensed under the MIT License.

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Latest release

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for Aquarids from gravatar.com Aquarids

License expression

MIT
View SPDX License List

Requires

Python >=3.10

Report project as malware

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

gpe_eval-0.2.0.tar.gz (11.7 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

gpe_eval-0.2.0-py3-none-any.whl (11.8 MB view details)

Uploaded Python 3

File details

Details for the file gpe_eval-0.2.0.tar.gz.

File metadata

  • Download URL: gpe_eval-0.2.0.tar.gz
  • Upload date:
  • Size: 11.7 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for gpe_eval-0.2.0.tar.gz
Algorithm Hash digest
SHA256 62439d3940796be169e76fb36b76ce470738dce35ec1a48f48881e895670cbad
MD5 597e024972b280b2fdb62fc310c3bbcb
BLAKE2b-256 b1f719421a2438326fd41173345096d3495205b6a95b5a62a1ff8119156c1091

See more details on using hashes here.

Provenance

The following attestation bundles were made for gpe_eval-0.2.0.tar.gz:

Publisher: pypi-publish.yml on Aquarids/GPE-pub

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file gpe_eval-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: gpe_eval-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 11.8 MB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for gpe_eval-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 91722c8719fe7c732332ab8e0320fac10f0f6d960a3a691510f4f0a9fa16d985
MD5 134ebb31ff90304ff27aac7de0ecb912
BLAKE2b-256 d434f1b1854ae596c99b4b70386507c497c96011e2399b3d8a4be757a34247b3

See more details on using hashes here.

Provenance

The following attestation bundles were made for gpe_eval-0.2.0-py3-none-any.whl:

Publisher: pypi-publish.yml on Aquarids/GPE-pub

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page