Log inSign up
Shiyi Cao
113 posts
Shiyi Cao profile banner
@shiyi_c98

Shiyi Cao

@shiyi_c98
llm and system | PhD student @UCBerkeley @BerkeleySky, MSc @ETH, B.S @sjtu1896 | Intern @nvidia
Berkeley, CA
shiyicao.com
Joined February 2019
796
Following
2,061
Followers
RepliesRepliesRepostsRepostsMediaMedia

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
  • Pinned
    @shiyi_c98
    Shiyi Cao
    @shiyi_c98
    Feb 26
    Introducing our new work K-Search: LLM Kernel Generation via Co-Evolving Intrinsic World Model — a new paradigm for automated GPU kernel generation, achieving SoTA results. 🔍 Big insight: Traditional methods treat LLMs as stochastic code generators inside heuristic loops — but
    Image
    Image
    12
  • @shiyi_c98
    Shiyi Cao
    @shiyi_c98
    Jul 9
    Biomni is now out in Science @ScienceMagazine ! Huge congrats to the team 🎉 Biomni is a really exciting step toward AI agents that can carry out real biomedical research workflows. Proud that SkyRL supported the RL training for Biomni-R0 early last year, where we explored
    @KexinHuang5
    Kexin Huang
    Phylo
    @KexinHuang5
    Jul 9
    Today, we're excited to share that Biomni is published in @ScienceMagazine. Biomedical research is still fragmented, manual, and difficult to scale. In this work, we introduce Biomni - the first general-purpose biomedical AI agent with an integrated biology environment that can
    Image
    5
  • @shiyi_c98
    Shiyi Cao
    @shiyi_c98
    May 27
    FlashLib🐮🐮🐮
    @Andy_ShuoYang
    Shuo Yang
    @Andy_ShuoYang
    May 27
    Flash-KMeans was only the beginning. Today, from the Flash-KMeans team, we are releasing FlashLib — a GPU library for fast, predictable, agent-ready classical ML operators. Up to 26× on KMeans, 19× on KNN, 40× on HDBSCAN, 208× on TruncatedSVD, 47× on PCA, 147× on exact t-SNE,
    Image
    00:00
  • @shiyi_c98
    Shiyi Cao
    @shiyi_c98
    Mar 6
    🤖🤖 Tried something fun today: asked Claude Code to create an agent team (an Implementer + a Planner) to implement the flashinfer mla paged decode CUDA kernel. The Implementer spent ~20 turns writing tests and debugging to use wgmma but kept getting stuck.😵‍💫😵‍💫😵‍💫 The Planner
    Image
    Image
    Image
    4
  • @shiyi_c98
    Shiyi Cao
    @shiyi_c98
    Dec 2, 2025
    🤖 I am in San Diego for #NeurIPS2025 this week! Excited to chat about SkyRL(-Agent), Coding LLM/Agent, Self-evolving Agent, RL, and Inference/Training Infrastructure.
    Image
    4
Advertisement
Advertisement