arxiv:2610.01509
🔄 In a Training Loop
Changdae Oh
changdae
AI & ML interests
Generalization; Distribution Shift; Uncertainty Quantification; Reward Modeling; Post-training
Recent Activity
upvoted a paper 2 days ago
Harness^2: Recursive Agent Harnessing for an Open World upvoted a paper 8 days ago
MIMESIS: Learning User Simulators as Training Environments for Interactive Agents upvoted a paper 8 days ago
On-Policy Distillation with Negative-Policy RolloutsOrganizations
datasets 7
changdae/tau2-uq-artifacts
Viewer • Updated • 2.25M • 672
changdae/vittle-pope-visual-perturbed
Updated • 110
changdae/vittle-llavabench-coco-textual-perturbed
Viewer • Updated • 30 • 81
changdae/vittle-llavabench-coco-joint-perturbed
Viewer • Updated • 90 • 64
changdae/vittle-llavabench-coco-visual-perturbed
Viewer • Updated • 270 • 131
changdae/llavabench-shift-natural-v1
Viewer • Updated • 1.47k • 57
changdae/llavabench-shift-synthetic-v1
Updated • 68