8 papers · 1 filter
RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents
Shuai Bai, Jiayong Deng, Sicheng Fan +30
Computer-use agents (CUAs) have advanced along two separate lines: graphical interaction and software development through code and the command line. Real digital work requires both…
Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments
Jie Wu, Zhenru Zhang, Beichen Zhang +11
As terminal-based code agents become prevalent, agent trajectories have accumulated at scale, while realistic, executable environments remain scarce. However, environments are what…
Gradients Know What Outcomes Don't: Unlocking Reinforcement Learning for LLM Reasoning with Gradient-Aligned Rewards
Leqi Zheng, Jinbo Su, Fang Niu +8
Reinforcement learning from verifiable rewards (RLVR) drives chain-of-thought reasoning in large language models, yet its binary outcome reward cannot distinguish among correct tra…
Learning Where Outcomes Change:Credit-Addressable Reasoning for Multimodal Geometry
Jiani Guo, Junjie Wang, Jie Wu +5
Multimodal geometry reasoning requires VLMs to extract precise visual relations and preserve them through multi-step deduction. Existing free-form traces obscure the decisions that…
DPIAgent: Divide, Protocol, Isolate for Agentic Reproduction Test Generation
Hao Liu, Steven Liu, Xin Zhang +8
Reproduction test generation, producing a failing-then-passing test that captures a reported bug, is a critical step in automated software engineering. Existing agentic methods tre…
VeriEvol: Scaling Multimodal Mathematical Reasoning via Verifiable Evol-Instruct
Haoling Li, Kai Zheng, Jie Wu +4
Scaling reinforcement learning for visual mathematical reasoning requires more than generating harder questions: as data volume grows, the reward labels themselves must remain reli…
ViPO: Visual Preference Optimization at Scale
Ming Li, Jie Wu, Justin Cui +3
While preference optimization is crucial for improving visual generative models, how to effectively scale this paradigm remains largely unexplored. Current open-source preference d…
X-Coder: Advancing Competitive Programming with Synthetic Tasks, Solutions, and Tests
Jie Wu, Haoling Li, Xin Zhang +8
Competitive programming remains challenging for code LLMs. Despite recent progress, many training pipelines still depend on scarce real-world data, raising concerns about scalabili…