ToDi: Token-wise Distillation via Fine-Grained Divergence Control
Paper • 2505.16297 • Published • 1
Seongryong Jung
SeongryongJung
AI & ML interests
Post-training, Knowledge Distillation, Agentic AI
Recent Activity
updated a dataset about 18 hours ago
SeongryongJung/multi-turn-rollout-chat liked a dataset 3 days ago
PrimeIntellect/INTELLECT-3-SFT liked a dataset 3 days ago
nvidia/Nemotron-Cascade-2-SFT-DataOrganizations
None yet
upvoted a paper 4 days ago