Loading activity…
@jiaxin_ge_
PhD @berkeley_ai ｜Visiting Researcher @AIatMeta ｜Prev: CS Undergrad @PKU1898 multimodal agent, omni model, agentic robotics
AI
A team member claims supervised fine-tuning lifted Agima’s harder-task result from 2% to 40%, versus GPT-5.6 Sol’s 20%.
XAI
The team says its year-long study examines when and how generation can help with understanding tasks.
X
