Hubert Strauss

PP_2.jpg

hs6702[at]princeton.edu

Hi! I am a Research Engineer at Princeton University, within the Princeton Language & Intelligence Initiative.

My current work focuses on reinforcement learning for language models, from studying how reward signals shape learning and optimization to building efficient systems and infrastructure for it. Since generation lies at the core of these systems, I am also interested in efficient generation methods.

I previously spent time at o9Solutions (ML team), and at Dematic (Operations Research and Applied AI team).

I obtained my engineering degree (BSc./MSc.) at Arts et Metiers ParisTech and studied optimization (MSc.) at Georgia Tech.

news

2026-Jan FutureFill has been accepted to ICLR 2026. See you in Rio !
2025-Aug Our work which provides an optimization perspective on what makes a good reward model for RLHF has been accepted to NeurIPS 2025. See you in San Jose !
2025-May Our work on Hardware-Efficient Attention for Fast Decoding has been accepted to COLM 2025. See you in Montreal !
2024-Jun I joined Princeton Language & Intelligence (Princeton University) as a Research Engineer

publications

  1. hardware_eff.png
    Hardware-Efficient Attention for Fast Decoding
    Ted Zadouri, Hubert Strauss, and Tri Dao
    In Conference on Language Modeling (COLM 2025), 2025
  2. what_makes_good_rm.png
    What Makes a Reward Model a Good Teacher? An Optimization Perspective
    Noam Razin, Zixuan Wang, Hubert Strauss, and 3 more authors
    In Advances in Neural Information Processing Systems (NeurIPS 2025), 2025
  3. futurefill.png
    FutureFill: Fast Generation from Convolutional Sequence Models
    Naman Agarwal, Xinyi Chen, Evan Dogariu, and 6 more authors
    In International Conference on Learning Representations (ICLR 2026), 2024