Scholar
Markus Peschl
Google Scholar ID: hinquVUAAAAJ
Qualcomm AI Research
Citations & Impact
Citations
133
H-index
3
i10-index
2
Publications
10
Co-authors
13
list available
Publications
Resume (English only)
Academic Achievements
- Published papers: 'From Code to Action: Hierarchical Learning of Diffusion-VLM Policies', 'Hybrid Training for Vision-Language-Action Models', 'Focusing on What Matters: Object-Agent-centric Tokenization for Vision Language Action models', 'Differentiable and Learnable Wireless Simulation with Geometric Transformers', 'NeuroSteiner: A Graph Transformer for Wirelength Estimation', 'Robust scheduling with GFlowNets', 'Learning Perturbations for Soft-Output Linear MIMO Demappers', 'MORAL: Aligning AI with Human Norms through Multi-Objective Reinforced Active Learning'.
Research Experience
- Senior deep learning researcher at Qualcomm AI Research; research areas cover hierarchical learning of diffusion-VLM policies, hybrid training for vision-language-action models, object-agent-centric tokenization for vision language action models, differentiable and learnable wireless simulation with geometric transformers, graph transformers for wirelength estimation in physical design, robust scheduling with GFlowNets, learning perturbations for soft-output linear MIMO demappers, and aligning AI with human norms through multi-objective reinforced active learning (MORAL).
Background
- A senior deep learning researcher dedicated to advancing AI research that is both theoretically rigorous and practically impactful. Research interests include reinforcement learning, the intersection of deep learning and combinatorial optimization, especially for wireless and chip design problems. Recently, his research focuses on generative models for sequence modeling, decision making, and embodied AI.
Miscellany
- In his free time, he enjoys playing the piano, practicing meditation, reading philosophy, and thinking about artificial general intelligence (AGI), preferably on the beach.
