NYU mathematics graduate and research intern at SAIS focused on self-supervised visual representation learning, JEPA world models, and planning
How media typically covers Yubo Wang
Yubo Wang as author
SpecForge team released SpecBundle Phase 1, a collection of production-grade EAGLE-3 model checkpoints and SpecForge v0.2 framework to address adoption gaps in speculative decoding for LLM inference acceleration.
“Author of "Speculative Decoding Models" in LMSYS”
Referenced in coverage
“Co-author on Google's ScientistTwo research about autonomous AI scientific discovery.”