08:00 - 08:15 coffee, setup
08:15 - 08:30 Introduction & Overview of CERC-AAI Research (Slides), Irina Rish (CERC-AAI Lab/UdeM/Mila)
08:30 - 08:45 From Geometry to Worlds and Actions: Repurposing Geometric Foundation Models for 3D World Generation and Policy Learning (Slides), Seungryong Kim (KAIST)
08:45 - 09:00 Feed-forward 3D reconstruction and novel view synthesis (Slides), Eunbyung Park (Yonsei University)
09:00 - 09:15 One Step Video Editing with One Step Diffusion Models (Slides), Gyeong-Moon Park (Korea University)
09:15- 09:25 coffee break
09:25 - 09:40 Advancing Data Attribution with Influence Functions (Slides), Dongwoo Kim (Postech)
09:40 - 09:55 Temporal Recurrence Favors Fewer Layers (Slides), Ivan Anokhin (Mila)
9:55 - 10:10 Learning Complex Behaviors from Offline Data (Slides), Jongmin Lee (Yonsei University)
10:10 - 10:25 On the Limits of Test-Time Compute for Self-Evaluation (Slides), Anirudh Buvanesh (Mila)
10:25 - 10:35 coffee break
10:35- 10:50 Target-Aware Direct Preference Optimization for Longitudinal Post-Operative Tumor MRI Generation (Slides), Tae-Eui Kam (Korea University)
10:50 - 11:0 Towards Interactive and Reasoning-Capable Time-Series Foundation Models (Slides), Changhee Lee (Korea University)
11:05 - 11:20 Characterizing Latent Reasoning Dynamics in LLMs, (Slides), Prateek Humane, (Mila)
11:20 - 11:35 Concluding Remarks