PPL
Haoran Zhang
Papers (5)
Draft-OPD: On-Policy Distillation for Speculative Draft ModelsAGORA: Adapter-Grounded Observation-Action Retention for Inference-Free Prompt Compression in LLM AgentsFrom I/O to Code with Discovery Agent$π$-Bench: Evaluating Proactive Personal Assistant Agents in Long-Horizon WorkflowsAchieving Gold-Medal-Level Olympiad Reasoning via Simple and Unified Scaling