LiveFact: A Dynamic, Time-Aware Benchmark for LLM-Driven Fake News Detection

1. 1Develop a dynamic benchmark that continuously updates with new data
2. 2Assess LLMs' ability to reason under temporal uncertainty
3. 3Evaluate the effectiveness of LiveFact in mitigating benchmark data contamination (BDC)

AI engineers and researchers working on LLMs and fake news detection tasks can benefit from LiveFact to evaluate and improve their models' performance over time

💡 Current static benchmarks are vulnerable to BDC and ineffective at assessing reasoning under temporal uncertainty

🚨 Introducing LiveFact: a dynamic benchmark for LLM-driven fake news detection under temporal uncertainty 🚨