3 papers
cs.AI2026
What Shared Prefixes Hide: Trajectory Dropout for On-Policy Distillation
Zizhuo Lin, Zzizhuo Lin, Quanling Liu +2
On-policy distillation (OPD) trains a student model on its own trajectories using dense token-level feedback from a stronger teacher model. Since each update is conditioned on the…
cs.CL2026
Same Evidence, Different Answers: Canonical-Context On-Policy Distillation for Multi-Turn Language Models
Zizhuo Lin, Quanling Liu, Jinsheng Quan +6
Large language models (LLMs) often solve a task when all instructions are given in a single prompt, but fail when the same information is revealed gradually across turns. When a cl…
cs.AI2024
COMET: "Cone of experience" enhanced large multimodal model for mathematical problem generation
Sannyuya Liu, Jintian Feng, Zongkai Yang +4
The automatic generation of high-quality mathematical problems is practically valuable in many educational scenarios. Large multimodal model provides a novel technical approach for…