Announcement_paper_neurips26

Five papers are accepted by NeurIPS 2026. One scales quantum bandits to over 1000 arms on IBM quantum hardware by tackling memory decoherence. The other attains the quantum speedup with only log*(n)+1 rounds of adaptivity. The other three cover multi-agent learning in matching markets, adversarial attacks on bandits, and fused reward and dueling feedback. Many thanks to all my collaborators!