Ran He
5 papers · Latest:
Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training
A new principle for LM post-training uses sparse rewards for strong teachers and dense distillation for students, outperforming direct sparse RL.
SpeechParaling-Bench: A Comprehensive Benchmark for Paralinguistic-Aware Speech Generation
SpeechParaling-Bench is a new benchmark for evaluating paralinguistic-aware speech generation in LALMs, using fine-grained features and a novel LALM-based judge.
Advancing Vision Transformer with Enhanced Spatial Priors
EVT enhances Vision Transformers by incorporating Euclidean spatial priors and flexible grouping, achieving state-of-the-art performance across vision tasks.
TIP: Token Importance in On-Policy Distillation
TIP introduces a two-axis taxonomy for token importance in on-policy distillation, significantly improving efficiency and reducing memory usage.
Are GUI Agents Focused Enough? Automated Distraction via Semantic-level UI Element Injection
This paper introduces Semantic-level UI Element Injection to red-team GUI agents by overlaying harmless UI elements, revealing model-agnostic vulnerabilities.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.