Log inSign up
Peter Schafhalter
118 posts
@pschafhalter

Peter Schafhalter

@pschafhalter
AI-Sys PhD student @ucbrise @ucberkeley focusing on systems for self-driving cars
pschafhalter.com
Joined May 2019
246
Following
364
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.
  • @pschafhalter
    Peter Schafhalter
    @pschafhalter
    Oct 29, 2024
    Introducing Modular Domain Experts (MoDE), a new multi-domain adaptation technique for LLMs. MoDE independently trains experts on different domains and composes them to boost LLM performance on complex, multi-domain tasks. Paper: arxiv.org/abs/2410.10181
    Image
    1
  • @pschafhalter
    Peter Schafhalter
    @pschafhalter
    Sep 24, 2024
    Great work by @justinywong_ @luo_michael1234! I've been having a lot of fun voting on the video models.
    @aivideoarena
    Video Arena
    @aivideoarena
    Sep 24, 2024
    🚀 Just Launched: VideoArena!🎥 Discover head-to-head comparisons of video clips generated from the same prompts across top text-to-video models. Compare outputs from 7 leading models and we're adding more soon! 🔗 Check out the leaderboard: videoarena.tv #Text2Video
    Image
    00:00
    1
  • @pschafhalter
    Peter Schafhalter
    @pschafhalter
    Aug 22, 2024
    Proud of what we came up with. The new Berkeley OS prelim balances breadth with depth by connecting key modern advances to ideas introduced in older papers. Syllabus: ucbosprelim.samkumar.org
    @shishirpatil_
    Shishir Patil
    @shishirpatil_
    Aug 21, 2024
    Replying to @conor_power23
    PSA: everybody just take the OS prelims. We rehauled the prelims last semester, and it’s now a pure riot 🎆🎉 cc @samkumar_cs @pschafhalter
  • @pschafhalter
    Peter Schafhalter
    @pschafhalter
    Aug 6, 2024
    Very cool to see another high-performance simulator for self-driving! GPUDrive addresses a key limitation of CARLA (simulation speed) and is even capable of simulating sensors like LiDARs.
    @EugeneVinitsky
    Eugene Vinitsky 🦋
    @EugeneVinitsky
    Aug 6, 2024
    We’re open-sourcing and arxiving GPUDrive, a GPU-accelerated 2.5D multi-agent driving simulator that runs at over a million FPS. Hundreds of scenes on one GPU means scalable multi-agent planning
    Image
    GIF
    1
  • @pschafhalter
    Peter Schafhalter
    @pschafhalter
    May 21, 2024
    Interesting approach from Waymo to use Q/A for remote interventions. Perhaps this is used to improve driving models via RLHF?
    @Waymo
    Waymo
    @Waymo
    May 21, 2024
    We’ve built the Waymo Driver to operate without the need for human intervention, and in order to do that sometimes, it requires additional context from our fleet response team. See their role in helping us safely scale: waymo.com/blog/2024/05/f…
    Image
    00:00
Advertisement
Advertisement