Re-implementation of the paper "Chain-of-Verification Reduces Hallucination in Large Language Models" for hallucination reduction. Developed as a final project of the Advanced Deep Learning course (DD3412) at KTH.

This repository contains the code of our paper 'Skip \n: A simple method to reduce hallucination in Large Vision-Language Models'.

Code & Data for our Paper "Alleviating Hallucinations of Large Language Models through Induced Hallucinations"

EMNLP'2024: Knowledge Verification to Nip Hallucination in the Bud

EMNLP'2024: Knowledge Verification to Nip Hallucination in the Bud

[ICLR'24] Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuning

Code for ACL 2024 paper "TruthX: Alleviating Hallucinations by Editing Large Language Models in Truthful Space"

This is the official repo for Debiasing Large Visual Language Models, including a Post-Hoc debias method and Visual Debias Decoding strategy.

Verify outputs generated by LLMs backed with real time data

[IJCAI 2024] FactCHD: Benchmarking Fact-Conflicting Hallucination Detection

openai assistant using code interpreter

openai function calling demo that gets customizable weather information

✨✨Woodpecker: Hallucination Correction for Multimodal Large Language Models. The first work to correct hallucinations in MLLMs.

Controlled HALlucination-Evaluation (CHALE) Question-Answering Dataset

Cocktail dynamic graph prompting technique in LLM for hallucination

[NAACL24] Official Implementation of Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information

OLAPH: Improving Factuality in Biomedical Long-form Question Answering

sFRC: To identify fakes in medical images reconstructed using AI

🧙🏻Code and benchmark for our Findings of ACL 2024 paper - "TimeChara: Evaluating Point-in-Time Character Hallucination of Role-Playing Large Language Models"