Papers

• THINKSAFE: Self-Generated Safety Alignment for Reasoning Models (2026)
  Seanie Lee et al.
  1.94
• Prompt-Level Distillation: A Non-Parametric Alternative to Model Fine-Tuning for Efficient Reasoning (2026)
  Sanket Badhe et al.
  1.94
• On-Policy Self-Distillation for Reasoning Compression (2026)
  Hejian Sang et al.
  1.94
• Self-Distillation Zero: Self-Revision Turns Binary Rewards into Dense Supervision (2026)
  Yinghui He et al.
  1.94
• Unmasking On-Policy Distillation: Where It Helps, Where It Hurts, and Why (2026)
  Mohammadreza Armandpour et al.
  1.94
• Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation (2026)
  Yuanyi Wang et al.
  1.94
• HYDRA-X: Native Unified Multimodal Models with Holistic Visual Tokenizers (2026)
  Guozhen Zhang et al.
  1.94
• Trust the Right Teacher: Quality-Aware Self-Distillation for GUI Grounding (2026)
  Jingyuan Huang et al.
  1.94
• Seeing Before Reasoning: Decoupling Perception and Reasoning for Shortcut-Resilient Multimodal On-Policy Self-Distillation (2026)
  Sihan Wang et al.
  1.94
• Trajectory as the Teacher: Few-Step Discrete Flow Matching via Energy-Navigated Distillation (2026)
  Amin Karimi Monsefi et al.
  1.89
• Nemotron-Cascade 2: Post-Training LLMs with Cascade RL and Multi-Domain On-Policy Distillation (2026)
  Zhuolin Yang et al.
  1.78
• OVD: On-policy Verbal Distillation (2026)
  Jing Xiong et al.
  1.67
• The Mirage of Model Editing: Revisiting Evaluation in the Wild (2025)
  Wanli Yang et al.
  1.28
• Overcoming Vocabulary Mismatch: Vocabulary-agnostic Teacher Guided Language Modeling (2025)
  Haebin Shin et al.
  1.28
• Distillation and Refinement of Reasoning in Small Language Models for Document Re-ranking (2025)
  Chris Samarinas et al.
  1.28
• Unleashing the Reasoning Potential of Pre-trained LLMs by Critique Fine-Tuning on One Problem (2025)
  Yubo Wang et al.
  1.28
• VA-π: Variational Policy Alignment for Pixel-Aware Autoregressive Generation (2025)
  Xinyao Liao et al.
  1.28
• On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes (2023)
  Rishabh Agarwal et al.
  —
• Distribution Backtracking Builds A Faster Convergence Trajectory for One-step Diffusion Distillation (2024)
  Shengyuan Zhang et al.
  —
• Iterative Graph Alignment (2024)
  Fangyuan Yu et al.
  —