<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Research on Chenyi Li</title>
    <link>https://www.chenyi-li.com/research_new/</link>
    <description>Recent content in Research on Chenyi Li</description>
    <image>
      <url>https://www.chenyi-li.com/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E</url>
      <link>https://www.chenyi-li.com/%3Clink%20or%20path%20of%20image%20for%20opengraph,%20twitter-cards%3E</link>
    </image>
    <generator>Hugo -- gohugo.io</generator>
    <language>en</language>
    <lastBuildDate>Fri, 30 Oct 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://www.chenyi-li.com/research_new/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Human-AI Synergy Supports Collective Creative Search</title>
      <link>https://www.chenyi-li.com/research_new/collective-creative-search/</link>
      <pubDate>Wed, 01 Jul 2026 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/collective-creative-search/</guid>
      <description>In a group word-guessing game, mixed teams of humans and AI found better answers than all-human or all-AI teams, and each side changed how the other searched.</description>
    </item>
    
    <item>
      <title>What Is the Role for Humans in Creative Writing? Creativity and Diversity in Human-AI Collaboration</title>
      <link>https://www.chenyi-li.com/research_new/creative-writing-humans-role/</link>
      <pubDate>Fri, 30 Oct 2026 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/creative-writing-humans-role/</guid>
      <description>AI writing chains produce creative but look-alike stories. Putting one human at the planning stage keeps the creativity and brings back the variety.</description>
    </item>
    
    <item>
      <title>Why Human Guidance Matters in Collaborative Vibe Coding</title>
      <link>https://www.chenyi-li.com/research_new/vibe-coding/</link>
      <pubDate>Fri, 06 Mar 2026 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/vibe-coding/</guid>
      <description>Across 20 experiments with 737 people, human instructions kept improving AI-written code while AI-written instructions made it get worse. The best split: humans instruct, AI evaluates.</description>
    </item>
    
    <item>
      <title>Satori: Towards Proactive AR Assistant with Belief-Desire-Intention User Modeling</title>
      <link>https://www.chenyi-li.com/research_new/satori/</link>
      <pubDate>Sat, 26 Apr 2025 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/satori/</guid>
      <description>An AR assistant that infers what you know, what you&amp;rsquo;re trying to do, and your next step, and then offers help before you ask. It matched assistants hand-designed by AR experts.</description>
    </item>
    
    <item>
      <title>Self-Supervised Learning for In Vivo Localization of Microelectrode Arrays Using Raw Local Field Potential</title>
      <link>https://www.chenyi-li.com/research_new/lfp2vec/</link>
      <pubDate>Thu, 04 Dec 2025 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/lfp2vec/</guid>
      <description>Lfp2vec adapts a speech model to raw brain recordings and learns to tell which brain region an electrode sits in, even for new labs, probe types, and species.</description>
    </item>
    
    <item>
      <title>Blind In Vivo Localization of Microelectrode Arrays via Functional Correlation Patterns in the Mouse Hippocampus</title>
      <link>https://www.chenyi-li.com/research_new/neuro-decode/</link>
      <pubDate>Sat, 05 Oct 2024 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/neuro-decode/</guid>
      <description>How channels on a 1024-channel probe correlate with each other is enough to tell which hippocampal layer each channel is in: 93% accuracy with only 5% of channels labeled.</description>
    </item>
    
    <item>
      <title>Embedding Decomposition for Artifacts Removal in EEG Signals</title>
      <link>https://www.chenyi-li.com/research_new/eeg-denoise/</link>
      <pubDate>Fri, 01 Apr 2022 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/eeg-denoise/</guid>
      <description>DeepSeparator removes eye-blink and muscle noise from EEG by separating signal from artifact inside a neural network&amp;rsquo;s embedding space, and shows you exactly what it removed.</description>
    </item>
    
    <item>
      <title>A Framework of Mapping Visual Stimulus to Neural Responses via Discrete Latent Representation</title>
      <link>https://www.chenyi-li.com/research_new/visual-neural-mapping/</link>
      <pubDate>Tue, 20 Feb 2024 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/visual-neural-mapping/</guid>
      <description>A V1 model that passes images through a small codebook of discrete features predicts neurons&amp;rsquo; responses as well as a standard CNN, as long as the readout is factorized into where and what.</description>
    </item>
    
    <item>
      <title>Learning to Reinforcement Learn with Transformer</title>
      <link>https://www.chenyi-li.com/research_new/meta-learn/</link>
      <pubDate>Thu, 01 Aug 2024 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/meta-learn/</guid>
      <description>A transformer trained only on random-walk experience learns, in context, to explore a new maze once and then head straight for the goal: learning to learn without updating its weights.</description>
    </item>
    
    <item>
      <title>Can a Smartwatch Strengthen Memory While You Sleep? Vibration-Cued Targeted Memory Reactivation at Home</title>
      <link>https://www.chenyi-li.com/research_new/tmr-smartwatch/</link>
      <pubDate>Sat, 31 May 2025 00:00:00 +0000</pubDate>
      
      <guid>https://www.chenyi-li.com/research_new/tmr-smartwatch/</guid>
      <description>A consumer smartwatch found deep sleep and delivered memory cues at home without waking people. The memory benefit was small and not reliable, and heavy cueing may have backfired.</description>
    </item>
    
  </channel>
</rss>
