Scholar
Salman Rahman
Google Scholar ID: vr7uTc8AAAAJ
University of California Los Angeles
Citations & Impact
Citations
447
H-index
9
i10-index
9
Publications
20
Co-authors
5
list available
Publications
Resume (English only)
Academic Achievements
- Publications: SPARK (reference-free RL training with generative process reward models), X-Teaming (adaptive multi-agent jailbreaks and defenses), MOSAIC (social AI for content dissemination), Xolver (multi-agent reasoning with holistic experience learning), AI Debate (scalable oversight for factuality claims). Several papers have been accepted at NeurIPS 2025, EMNLP 2025, and COLM 2025.
Research Experience
- Interned with Amazon's AGI team, working on generative process reward models for improving LLM reasoning through reinforcement learning; Interned at Apple's machine learning team, developing efficient multimodal LMs for on-device deployment; Helps organize the NLP Seminar Series at UCLA; Before joining UCLA, worked on projects including Clinical LLM generalization, machine learning explanation disparity, and big data in healthcare at NYU.
Education
- Ph.D. Student in Computer Science at the University of California, Los Angeles (UCLA), supervised by Professor Saadia Gabriel; Collaborates closely with Professors Pavel Izmailov and Yejin Choi.
Background
- Research Interests: Improving the reasoning and planning capabilities of language models through reinforcement learning; Field: AI/NLP; Brief Introduction: Focused on developing AI systems that can perform multi-step reasoning, devise and execute plans, use tools effectively, and collaborate through communication and debate.
Miscellany
- During undergraduate and master's studies, focused on computational sustainability, exploring how AI systems can address pressing societal and environmental challenges of the 21st century.
