RZ

Runsong Zhao

cs.CLcs.AIcs.LGcs.SDeess.AS

On Valency

published · living versions

W_af3g4cpf·v1 · currentpublished

PEAR: Adaptive Advantage Reshaping for Perception and Exploration in Large Vision-Language Models

with Junxin Wang, Yingfeng Luo, Xiangchen Song, Weiqiao Shan +10

1 version

Preprints & journals

9 papers in the corpus · 2025–2026

Complementary Roles of Activation and Parametric Memory in Few-Shot Learning2609.28250v1 · Miaohe Niu, Runsong Zhao, Xinyu Liu et al.2026 · 0 citationsarXiv

HybridThinker: Efficient Chain-of-Thought Reasoning via Compressed Memory and Transient Thought Steps2606.03768v1 · Xin Liu, Runsong Zhao, Xinyu Liu et al.2026 · 0 citationsarXiv

Teacher-Guided Policy Optimization for On-Policy Reasoning Distillation under Large Policy Divergence2605.13230v2 · Xinyu Liu, Kechen Jiao, Chunyang Xiao et al.2026 · 0 citationsarXiv

MemoSight: Unifying Context Compression and Multi Token Prediction for Reasoning Acceleration2604.14889v2 · Xinyu Liu, Xin Liu, Bo Jin et al.2026 · 0 citationsarXiv

PoC: Performance-oriented Context Compression for Large Language Models via Performance Prediction2603.19733v1 · Runsong Zhao, Shilei Liu, Jiwei Tang et al.2026 · 0 citationsarXiv

CoMeT: Collaborative Memory Transformer for Efficient Long Context Modeling2602.01766v2 · Runsong Zhao, Shilei Liu, Jiwei Tang et al.2026 · 0 citationsarXiv

Read As Human: Compressing Context via Parallelizable Close Reading and Skimming2602.01840v2 · Jiwei Tang, Shilei Liu, Zhicheng Zhang et al.2026 · 0 citationsarXiv

MTP-S2UT: Enhancing Speech-to-Speech Translation Quality with Multi-token Prediction2510.10003v1 · Jianjin Wang, Runsong Zhao, Xiaoqian Liu et al.2025 · 0 citationsarXiv

Autoencoding-Free Context Compression for LLMs via Contextual Semantic Anchors2510.08907v4 · Xin Liu, Runsong Zhao, Pengcheng Huang et al.2025 · 0 citationsarXiv

Career total: 12 works. 9 are in this corpus.

Profile built from the corpus for this byline.

Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.