Log inSign up
user avatar
Eunsol Choi
@eunsolc
on natural language processing / machine learning. assistant prof at @NYUDataScience @NYU_Courant prev @UTCompSci @googleai, @uwcse, @Cornell.
eunsol.github.io
Joined September 2016
976
Following
6,467
Followers
RepliesMedia
  • Pinned
    user avatar
    Eunsol Choi
    @eunsolc
    Aug 16, 2024
    My lab will move to @NYUDataScience and @NYU_Courant this Fall! I’m excited to connect with amazing researchers at @CILVRatNYU and larger ML/NLP community in NYC. I will be recruiting students this cycle at NYU. Happy to be back to the city 🗽on the east coast as well. I had a
  • user avatar
    Eunsol Choi
    @eunsolc
    Apr 29
    We study sampling diverse output from a suite of LLMs. One key surprise for me was that it's better to carefully pick a single model to sample many times, rather than naively mixing outputs from multiple models.
    user avatar
    Yuhan Liu
    @YuhanLiu_nlp
    Apr 28
    Can LLMs generate diverse outputs for open-ended questions? Is it helpful if we ensemble outputs from multiple models? We study 18 LLMs on 4 datasets and find that no single model is best at generating diverse outputs 👇/ 🧵
    Image
  • user avatar
    Eunsol Choi
    @eunsolc
    Apr 9
    Do LLMs suffer from human-like cognitive biases? 🤔 Check out @arhjhaveri's new paper on how models navigate hypothesis spaces. We found that confirmation bias degrades LLM performance, and we explore strategies to mitigate it.
    user avatar
    Ayush Jhaveri
    @arhjhaveri
    Apr 8
    Your AI Agent just formed a hypothesis. 💭 How does it validate it? Not by trying to prove itself wrong. Rather, it selectively seeks evidence that confirms what it already believes, often ending up with the wrong answer! Confirmation bias isn’t just human. We measure it in
    Image
  • user avatar
    Eunsol Choi
    @eunsolc
    Feb 24
    Excited to introduce our new work on improving information coverage for search systems (LLMs + retriever). Key idea: train retrievers to search for answers that are missing from earlier retrieval rounds.
    user avatar
    Deniz Qian
    @denq1an
    Feb 24
    🚨NEW PAPER🚨 How can we comprehensively retrieve all relevant docs for multi-answer QA? Agentic search doesn't help. Introducing RVR, an iterative framework that conditions on prior docs to maximize answer coverage. 📈10% answer recall gain on QAMPARI w/@hungting_chen @eunsolc
    Image
    Image
  • user avatar
    Eunsol Choi
    @eunsolc
    Nov 4, 2025
    Reasoning models are more uncertain when generating reasoning traces for easy questions compared to generating traces for medium-difficulty questions. We use this insight to build a light-weight, adaptive inference strategy for reasoning models. Was fun exploring this with my
    user avatar
    Xiang Liu
    @Dominicliu12
    Nov 3, 2025
    LLMs spend too much time thinking... even on easy problems. 🤔 We introduce DiffAdapt, a lightweight framework for token-efficient reasoning that dynamically adapts to problem difficulty. 💡 It achieves 11.2% performance gain while reducing token usage by up to 22.4% and 6x
    Image