Shahar Levy

What if you could automatically turn a large collection of documents into structured databases, tailored for your own research needs?📄 We introduce ScheMatiQ! From question ➝ schema ➝ structured data 🔍 @ShaharLevy19 @MintzReshef @RKeydar @BarakRaveh @GabiStanovsky 🧵 👇1/5

Our results indicate that processing multiple documents is a separate challenge from handling long contexts. Joint Work with @NirMMazor @LihiShalmon @MichaelHassid @GabiStanovsky 3/3 @nlphuji

Previous work shows that adding more documents can hurt LLM performance, but didn’t isolate the impact of document quantity under a fixed context length. Our findings reveal that increasing the number of retrieved documents in RAG settings significantly challenges LLMs.🤯 2/3

Can RAG performance get * worse * with more relevant documents?📄 We put the number of retrieved documents in RAG to the test! 💥Preprint💥: arxiv.org/abs/2503.04388 1/3

📊📈🎯Happy to share a new benchmark: SEAM 🤐 - A Stochastic Evaluation Approach for Multi-document tasks Paper arxiv.org/pdf/2406.16086 Website seam-benchmark.github.io Code github.com/seam-benchmark… w/ @clu_avi @ArieCattan @ShaharLevy19 @obspp18 @GabiStanovsky