ZX
Zhiheng Xi
cs.CLcs.AIcs.LGcs.CVcs.SEcs.IRcs.CCcs.CRcs.ROcs.SD
On Valency
published · living versionsW_arrft6hu·v1 · currentpublished
ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents
with Guanting Dong, Zhichao Sun, Xinyu Yang, Rui Wang +20
1 version
Preprints & journals
83 papers in the corpus · 2022–2026QREF: Improving DeepSearch Agents by Aligning Search with Reasoning via Query Refinement and Evidence Filtering2610.21512v1 · Yujia Chen, Yuhui Wang, Junzhe Wang et al.2026 · 0 citationsarXiv
WorkRover: Scaling Long-Horizon Cowork Agent Training Via Recursive Environment Evolution2610.18067v1 · Hang He, Linting Wang, Guanming Liu et al.2026 · 0 citationsarXiv
Scaling Verifiable Environments for Long-horizon Work Agents2610.04906v5 · Jiazheng Zhang, Long Ma, Yunxian Yang et al.2026 · 0 citationsarXiv
ReDraft, Don't Just Distill: Reference-Driven Revision for Continual VLLM Post-Training2609.16639v3 · Zhihao Zhang, Mingqi Wu, Qiaole Dong et al.2026 · 0 citationsarXiv
Sci-MMR: Benchmarking Multi-Step Evidence-Grounded Scientific Reasoning in Multimodal Agents2609.11243v2 · Jiaqiang Li, Yajie Yang, Zhiheng Xi et al.2026 · 0 citationsarXiv
A Token-Level Analysis of Sampled-Token Reverse-KL On-Policy Distillation2608.25643v3 · Bing Shao, Jiazheng Zhang, Long Ma et al.2026 · 0 citationsarXiv
CAFE: Self-Improving Search Agents Need Co-Evolving Feedback2608.24794v2 · Boyang Liu, Senjie Jin, Peixin Wang et al.2026 · 0 citationsarXiv
An AI System for Autonomous Algorithm Evolution in Drug Development10.64898/2026.08.16.745117v1 · Zhou, Z., Nan, Y., Mou, M. et al.2026 · 0 citationsbioRxiv
IACM-RL: Intent-Aware Context Management and Reinforcement Learning for Complex Tool Invocation under Dynamic Intent Fluctuations2608.02110v2 · Dingwei Zhu, Jiahan Li, Chengjun Pan et al.2026 · 0 citationsarXiv
NexForge: Scaling Agent Capabilities through Requirement-Driven Task Synthesis for LLMs2607.14186v6 · Jiarong Zhao, Zhikai Lei, Zhiheng Xi et al.2026 · 0 citationsarXiv
AgentGym2: Benchmarking Large Language Model Agents in De-Idealized Real-World Environments2607.05174v1 · Zhiheng Xi, Dingwen Yang, Jiaqi Liu et al.2026 · 0 citationsarXiv
AGORA: An Archive-Grounded Benchmark for Agentic Workplace Document Reasoning2606.24526v2 · Honglin Guo, Qi Zhang, Yu Zhang et al.2026 · 0 citationsarXiv
Entropy Is Not Enough: Unlocking Effective Reinforcement Learning for Visual Reasoning via Vision-Anchored Token Selection2606.03937v2 · Senjie Jin, Peixin Wang, Boyang Liu et al.2026 · 0 citationsarXiv
AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security2605.29801v1 · Dongrui Liu, Yu Li, Zhonghao Yang et al.2026 · 0 citationsarXiv
HyperLogic: A Hard, Forward-Authored Chinese Logical Reasoning Benchmark with Execution-Derived Answers2605.19597v2 · Ming Zhang, Qiyuan Peng, Yinxi Wei et al.2026 · 0 citationsarXiv
Entropy Polarity in Reinforcement Fine-Tuning: Direction, Asymmetry, and Control2605.11775v2 · Jiazheng Zhang, Ziche Fu, Junrui Shen et al.2026 · 0 citationsarXiv
Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses2604.25850v4 · Jiahang Lin, Shichun Liu, Chengjun Pan et al.2026 · 1 citationarXiv
AgentV-RL: Scaling Reward Modeling with Agentic Verifier2604.16004v1 · Jiazheng Zhang, Ziche Fu, Zhiheng Xi et al.2026 · 0 citationsarXiv
MM-Doc-R1: Training Agents for Long Document Visual Question Answering through Multi-turn Reinforcement Learning2604.13579v1 · Jiahang Lin, Kai Hu, Binghai Wang et al.2026 · 0 citationsarXiv
Enhancing LLM-based Search Agents via Contribution Weighted Group Relative Policy Optimization2604.14267v2 · Junzhe Wang, Zhiheng Xi, Yajie Yang et al.2026 · 0 citationsarXiv
JFTA-Bench: Evaluate LLM's Ability of Tracking and Analyzing Malfunctions Using Fault Trees2603.22978v2 · Yuhui Wang, Zhixiong Yang, Ming Zhang et al.2026 · 0 citationsarXiv
AI Can Learn Scientific Taste2603.14473v3 · Jingqi Tong, Mingzhe Li, Hangcheng Li et al.2026 · 0 citationsarXiv
Can RL Improve Generalization of LLM Agents? An Empirical Study2603.12011v1 · Zhiheng Xi, Xin Guo, Jiaqi Liu et al.2026 · 0 citationsarXiv
SciAgentGym: Benchmarking Multi-Step Scientific Tool-use in LLM Agents2602.12984v2 · Yujiong Shen, Yajie Yang, Zhiheng Xi et al.2026 · 1 citationarXiv
DFPO: Scaling Value Modeling via Distributional Flow towards Robust and Generalizable LLM Post-Training2602.05890v2 · Dingwei Zhu, Zhiheng Xi, Shihan Dou et al.2026 · 0 citationsarXiv
Steering LLMs via Scalable Interactive Oversight2602.04210v2 · Enyu Zhou, Zhiheng Xi, Long Ma et al.2026 · 0 citationsarXiv
ChartE: A Comprehensive Benchmark for End-to-End Chart Editing2601.21694v1 · Shuo Li, Jiajun Sun, Zhekai Wang et al.2026 · 0 citationsarXiv
AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and Security2601.18491v2 · Dongrui Liu, Qihan Ren, Chen Qian et al.2026 · 0 citationsarXiv
Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models2601.14004v4 · Hengyuan Zhang, Zhihao Zhang, Mingyang Wang et al.2026 · 2 citationsarXiv
Which Reasoning Trajectories Teach Students to Reason Better? A Simple Metric of Informative Alignment2601.14249v5 · Yuming Yang, Mingyoung Lai, Wanxu Zhao et al.2026 · 0 citationsarXiv
FRoM-W1: Towards General Humanoid Whole-Body Control with Language Instructions2601.12799v1 · Peng Li, Zihan Zhuang, Yangfan Gao et al.2026 · 0 citationsarXiv
ABC-Bench: Benchmarking Agentic Backend Coding in Real-World Development2601.11077v1 · Jie Yang, Honglin Guo, Li Ji et al.2026 · 0 citationsarXiv
OpenNovelty: An LLM-powered Agentic System for Verifiable Scholarly Novelty Assessment2601.01576v2 · Ming Zhang, Kexin Tan, Yueyuan Huang et al.2026 · 1 citationarXiv
Memory in the Age of AI Agents2512.13564v2 · Yuyang Hu, Shichun Liu, Yanwei Yue et al.2025 · 1 citationarXiv
Nex-N1: Agentic Models Trained via a Unified Ecosystem for Large-Scale Environment Construction2512.04987v1 · Nex-AGI Team: Yuxuan Cai, Lu Chen, Qiaoling Chen et al.2025 · 0 citationsarXiv
DVPO: Distributional Value Modeling-based Policy Optimization for LLM Post-Training2512.03847v3 · Dingwei Zhu, Zhiheng Xi, Shihan Dou et al.2025 · 0 citationsarXiv
AI Deception: Risks, Dynamics, and Controls2511.22619v2 · Boyuan Chen, Sitong Fang, Jiaming Ji et al.2025 · 0 citationsarXiv
AgentPRM: Process Reward Models for LLM Agents via Step-Wise Promise and Progress2511.08325v1 · Zhiheng Xi, Chenyang Liao, Guanyu Li et al.2025 · 4 citationsarXiv
Parrot: A Training Pipeline Enhances Both Program CoT and Natural Language CoT for Reasoning2510.25310v1 · Senjie Jin, Lu Chen, Zhiheng Xi et al.2025 · 0 citationsarXiv
BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping2510.18927v1 · Zhiheng Xi, Xin Guo, Yang Nan et al.2025 · 0 citationsarXiv
From Scores to Preferences: Redefining MOS Benchmarking for Speech Quality Reward Modeling2510.00743v1 · Yifei Cao, Changhao Jiang, Jiabao Zhuang et al.2025 · 0 citationsarXiv
AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learning2509.08755v1 · Zhiheng Xi, Jixuan Huang, Chenyang Liao et al.2025 · 0 citationsarXiv
Feedback-Driven Tool-Use Improvements in Large Language Models via Automated Build Environments2508.08791v3 · Junjie Ye, Changhao Jiang, Zhengyin Du et al.2025 · 1 citationarXiv
LLMEval-Fair: A Large-Scale Longitudinal Study on Robust and Fair Evaluation of Large Language Models2508.05452v7 · Ming Zhang, Yujiong Shen, Jingyi Deng et al.2025 · 0 citationsarXiv
VRPO: Rethinking Value Modeling for Robust RL under Noisy Supervision in LLM Post-Training2508.03058v2 · Dingwei Zhu, Shihan Dou, Zhiheng Xi et al.2025 · 0 citationsarXiv
Pre-Trained Policy Discriminators are General Reward Models2507.05197v2 · Shihan Dou, Shichun Liu, Yuming Yang et al.2025 · 0 citationsarXiv
BMMR: A Large-Scale Bilingual Multimodal Multi-Discipline Reasoning Dataset2507.03483v2 · Zhiheng Xi, Guanyu Li, Yutao Fan et al.2025 · 0 citationsarXiv
LLMEval-Med: A Real-world Clinical Benchmark for Medical LLMs with Physician Validation2506.04078v3 · Ming Zhang, Yujiong Shen, Zelin Li et al.2025 · 7 citationsarXiv
Game-RL: Synthesizing Multimodal Verifiable Game Data to Boost VLMs' General Reasoning2505.13886v8 · Jingqi Tong, Jixin Tang, Hangcheng Li et al.2025 · 0 citationsarXiv
Two Minds Better Than One: Collaborative Reward Modeling for LLM Alignment2505.10597v3 · Jiazheng Zhang, Wenqing Jing, Zizhuo Zhang et al.2025 · 0 citationsarXiv
Career total: 95 works. 83 are in this corpus.Showing the 50 most recent.
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.