PPaperPicks

Martin Jaggi

EPFL, School of Computer and Communication Sciences, Lausanne, Switzerland

24 papers at tracked venues · 23 at CORE A* · active 2024–2026

DBLP profile ↗ORCID 0000-0003-1579-5558 ↗Google Scholar ↗Homepage ↗

Venues

NeurIPS ×9ICML ×7ICLR ×5AAAI ×1ACL ×1AISTATS ×1

Frequent coauthors

Atli Kosson
DBLP profile ↗ORCID search ↗
×3
Alejandro Hernández-Cano
DBLP profile ↗ORCID search ↗
×2
Ashok Vardhan Makkuva
DBLP profile ↗ORCID search ↗
×2
Simin Fan
DBLP profile ↗ORCID search ↗
×2
Dongyang Fan
DBLP profile ↗ORCID search ↗
×2
Amirkeivan Mohtashami
DBLP profile ↗ORCID search ↗
×1
Simla Burcu Harma
DBLP profile ↗ORCID search ↗
×1
Bettina Messmer
DBLP profile ↗ORCID search ↗
×1
El Mahdi Chayti
DBLP profile ↗ORCID search ↗
×1
Vinitra Swamy
DBLP profile ↗ORCID search ↗
×1
Diba Hashemi
DBLP profile ↗ORCID search ↗
×1
Matteo Pagliardini
DBLP profile ↗ORCID search ↗
×1

Papers

  1. Apertus: Democratizing Open and Compliant LLMs for Global Language Environments
    ACL 2026 ·
    Alejandro Hernández-Cano
    DBLP profile ↗ORCID search ↗
  2. Attention with Markov: A Curious Case of Single-layer Transformers
    ICLR 2025 ·
    Ashok Vardhan Makkuva
    DBLP profile ↗ORCID search ↗
  3. CoTFormer: A Chain of Thought Driven Architecture with Budget-Adaptive Computation Cost at Inference
    ICLR 2025 ·
    Amirkeivan Mohtashami
    DBLP profile ↗ORCID search ↗
  4. Effective Interplay between Sparsity and Quantization: From Theory to Practice
    ICLR 2025 ·
    Simla Burcu Harma
    DBLP profile ↗ORCID search ↗
  5. Enhancing Multilingual LLM Pretraining with Model-Based Data Selection
    NeurIPS 2025 ·
    Bettina Messmer
    DBLP profile ↗ORCID search ↗
  6. GRAPE: Optimize Data Mixture for Group Robust Multi-target Adaptive Pretraining
    NeurIPS 2025 ·
    Simin Fan
    DBLP profile ↗ORCID search ↗
  7. Improving Stochastic Cubic Newton with Momentum
    AISTATS 2025 ·
    El Mahdi Chayti
    DBLP profile ↗ORCID search ↗
  8. Intrinsic User-Centric Interpretability through Global Mixture of Experts
    ICLR 2025 ·
    Vinitra Swamy
    DBLP profile ↗ORCID search ↗
  9. On-Device Collaborative Language Modeling via a Mixture of Generalists and Specialists
    ICML 2025 ·
    Dongyang Fan
    DBLP profile ↗ORCID search ↗
  10. Towards Fully FP8 GEMM LLM Training at Scale
    NeurIPS 2025 ·
    Alejandro Hernández-Cano
    DBLP profile ↗ORCID search ↗
  11. URLs Help, Topics Guide: Understanding Metadata Utility in LLM Training
    NeurIPS 2025 ·
    Dongyang Fan
    DBLP profile ↗ORCID search ↗
  12. Analyzing & Reducing the Need for Learning Rate Warmup in GPT Training
    NeurIPS 2024 ·
    Atli Kosson
    DBLP profile ↗ORCID search ↗
  13. CoBo: Collaborative Learning via Bilevel Optimization
    NeurIPS 2024 ·
    Diba Hashemi
    DBLP profile ↗ORCID search ↗
  14. DOGE: Domain Reweighting with Generalization Estimation
    ICML 2024 ·
    Simin Fan
    DBLP profile ↗ORCID search ↗
  15. DenseFormer: Enhancing Information Flow in Transformers via Depth Weighted Averaging
    NeurIPS 2024 ·
    Matteo Pagliardini
    DBLP profile ↗ORCID search ↗
  16. Ghost Noise for Regularizing Deep Neural Networks
    AAAI 2024 ·
    Atli Kosson
    DBLP profile ↗ORCID search ↗
  17. LASER: Linear Compression in Wireless Distributed Optimization
    ICML 2024 ·
    Ashok Vardhan Makkuva
    DBLP profile ↗ORCID search ↗
  18. Layer-wise linear mode connectivity
    ICLR 2024 ·
    Linara Adilova
    DBLP profile ↗ORCID search ↗
  19. On Convergence of Incremental Gradient for Non-convex Smooth Functions
    ICML 2024 ·
    Anastasia Koloskova
    DBLP profile ↗ORCID search ↗
  20. QuaRot: Outlier-Free 4-Bit Inference in Rotated LLMs
    NeurIPS 2024 ·
    Saleh Ashkboos
    DBLP profile ↗ORCID search ↗
  21. Rotational Equilibrium: How Weight Decay Balances Learning Across Neural Networks
    ICML 2024 ·
    Atli Kosson
    DBLP profile ↗ORCID search ↗
  22. Scaling Laws and Compute-Optimal Training Beyond Fixed Training Durations
    NeurIPS 2024 ·
    Alexander Hägele
    DBLP profile ↗ORCID search ↗
  23. Spectral Preconditioning for Gradient Methods on Graded Non-convex Functions
    ICML 2024 ·
    Nikita Doikov
    DBLP profile ↗ORCID search ↗
  24. The Privacy Power of Correlated Noise in Decentralized Learning
    ICML 2024 ·
    Youssef Allouah
    DBLP profile ↗ORCID search ↗