LogicCheckGPT

Logic Checker

An open-source framework for detecting and mitigating object hallucinations in large vision-language models by analyzing logical consistency.

[ACL 2024] Logical Closed Loop: Uncovering Object Hallucinations in Large Vision-Language Models. Detect and mitigate object hallucinations in LVLMs by itself through logical closed loops.

GitHub

19 stars
4 watching
1 forks
Language: Python
last commit: over 2 years ago

Related projects:

RepositoryDescriptionStars
openkg-org/easydetectA framework to detect and mitigate hallucinations in multimodal large language models48
awakesecurity/spectacleA tool for writing and validating temporal logic specifications of software behavior174
amazon-science/refcheckerAutomates fine-grained hallucination detection in large language model outputs325
bradyfu/woodpeckerA method to correct hallucinations in multimodal large language models without requiring retraining617
melevir/flake8-cognitive-complexityA tool to measure and enforce the complexity of cognitive functions in Python code69
s-weigand/flake8-nbA tool to check Python code quality in Jupyter notebooks.28
bcdnlp/faithscoreEvaluates answers generated by large vision-language models to assess hallucinations27
jsh9/pydoclintA tool to check Python docstrings for consistency with function signatures or implementations.159
fluidex/plonkitProvides tools and utilities for generating and verifying proofs in a zkSNARK proof system159
thesheeep/godotccdA module that provides fast collision checking for arbitrary 3D shapes.64
crocmagnon/fatcontextDetects potential performance issues in Go code caused by nested contexts in loops or function literals13
damo-nlp-sg/vcdAn approach to reduce object hallucinations in large vision-language models by contrasting output distributions derived from original and distorted visual inputs222
umbrelladocs/linkspectorA CLI tool that checks for dead hyperlinks in files using multiple markup languages.70
fkie-cad/cwe_checkerA tool for detecting common bugs in binary executables.1,155
lalbj/paiImproves the performance of large language models by intervening in their internal workings to reduce hallucinations83