ARISE‑RL is a full‑cycle self‑evolution framework that couples a task/rubric generator with a reasoning solver, enabling open‑ended agents to learn via reinforcement learning without relying on gold answers. The generator creates tool‑grounded rubric criteria and rewards itself for producing valid, intermediate‑difficulty tasks, while the solver improves through fine‑grained rubric satisfaction signals and multi‑step reasoning. The framework also introduces Reward‑Gated Self‑Evolution Distillation to selectively distill memory‑augmented policies, reducing distribution mismatch, and it is evaluated on the new ECR‑Bench rubric suite, achieving state‑of‑the‑art performance across all benchmarks.
By Fanrui Zhang, Ruixue Ding, Qiang Zhang, Xi Chen, Boli Chen, Shihang Wang, Qiuchen Wang, Hongmin Zhan, Jinxin Bian, Li xingchao, Peijin Zheng, Hao cheng, Pengjun Xie, Kaipeng Zhang, Jiawei Liu, Zheng-Jun Zha
arXiv:2510. 05592v2 Announce Type: replace Abstract: Outcome-driven reinforcement learning has advanced reasoning in large language models (LLMs), but prevailing tool-augmented approaches train a single, monolithic policy that interleaves thoughts and tool calls under full context; this scales poorly with long horizons and diverse tools and generalizes weakly to new scenarios.
By Zhuofeng Li, Haoxiang Zhang, Seungju Han, Sheng Liu, Jianwen Xie, Yu Zhang, Yejin Choi, James Zou, Pan Lu
Behavior2Trip introduces a new task—Behavior‑Aware Travel Planning—where user preferences are inferred directly from past behavior trajectories rather than explicit instructions. The benchmark contains 11,400 Chinese travel‑planning instances, each with an average of 39.8 past behaviors across 14 attributes and 5 preference dimensions. A reinforcement‑learning agent, B2T‑Agent, leveraging behavior trajectories, external retrieval tools, and internal memory, outperforms strong baselines such as GPT‑4.1 on this challenging dataset.
arXiv:2601. 06487v3 Announce Type: replace-cross Abstract: Reinforcement learning has substantially improved the performance of LLM agents on tasks with verifiable outcomes, but it still struggles on open-ended agent tasks with vast solution spaces (e.
By Qiang Zhang, Boli Chen, Fanrui Zhang, Ruixue Ding, Shihang Wang, Qiuchen Wang, Yinfeng Huang, Haonan Zhang, Rongxiang Zhu, Pengyong Wang, Ailin Ren, Xin Li, Pengjun Xie, Jiawei Liu, Ning Guo, Jingren Zhou, Zheng-Jun Zha
Behavior2Trip introduces a new task—Behavior‑Aware Travel Planning—where user preferences are inferred from past behavior trajectories rather than explicit instructions. The benchmark contains 11,400 instances from a major Chinese travel platform, each with nearly 40 recorded behaviors across 14 attributes and 5 preference dimensions. A reinforcement‑learning agent, B2T‑Agent, leverages these trajectories, external retrieval tools, and internal memory, outperforming GPT‑4.1 and other baselines on the dataset.
By Zihao Cheng, Yingyu Shan, Hongru Wang, Zeming Liu, Xinyi Wang, Xiangrong Zhu, Yuhang Guo, Wei Lin, Yunhong Wang
arXiv:2608.30924v1 Announce Type: new
Abstract: Travel itinerary generation requires balancing strict spatio-temporal constraints with human preferences. Existing LLM-based planners mainly rely on st...
By Priyanshu Karmakar, Borru Vijay Sai, Shubhojit Mallick, Abhik Jana, Shreya Ghosh, Manish Gupta
arXiv:2608. 04588v1 Announce Type: cross Abstract: Agentic systems have emerged as a promising paradigm for solving complex tasks by coordinating specialized LLM-based agents.
By Junnan Liu, Linhao Luo, Thuy-Trang Vu, Gholamreza Haffari
arXiv:2606. 01046v1 Announce Type: new Abstract: The development of Large Language Models (LLMs) has significantly improved travel planning applications, yet evaluating such models is limited by existing benchmarks' limitations: 1) overemphasis on constraint compliance, neglecting multi-dimensional qualities like spatio-temporal cost; 2) datasets lacking real-world authenticity and coverage in key areas (e.
By Weiyi Chen, Shuaixiong Wang, Ziyun Gao, Kaichun Hu, Wangze Ni, Shimin Di, Chen Jason Zhang, Lei Chen
UnifiedPlayers is a cooperative framework that jointly adapts planning, execution, and evaluation for tool-integrated reinforcement learning agents. It consists of a Planning Player that generates tasks, an Execution Player that creates multi-turn trajectories with Python tool calls, and an Evaluation Player that builds executable verifiers, all coordinated by role‑specific rewards under GRPO. The approach outperforms prior baselines on mathematical and general reasoning benchmarks and yields a verifier with high adversarial detection accuracy and more discriminative reward signals.
By Wenjie Liao, Liangjie Zhao, Zehong Cao
arXiv:2512. 03438v3 Announce Type: replace Abstract: Agentic reasoning models trained with multimodal reinforcement learning (MMRL) have become increasingly capable, yet they are almost universally optimized using sparse, outcome-based rewards computed based on the final answers.
By Reuben Tan, Baolin Peng, Zhengyuan Yang, Hao Cheng, Oier Mees, Theodore Zhao, Andrea Tupini, Isar Meijer, Qianhui Wu, Yuncong Yang, Lars Liden, Yu Gu, Sheng Zhang, Xiaodong Liu, Lijuan Wang, Marc Pollefeys, Yong Jae Lee, Jianfeng Gao
TripScore is a benchmark and evaluation framework for large language models (LLMs) in travel planning, built from real user logs and calibrated with 1,468 pairwise judgments from 203 travel experts. It uses a hierarchical feasibility gate for format and commonsense checks, and a unified point-wise reward that combines soft quality and preference fulfillment. Experiments show that reinforcement learning fine‑tuning, such as GRPO, consistently outperforms other methods when evaluated with TripScore.
By Yincen Qu, Huan Xiao, Feng Li, Gregory Li, Hui Zhou, Xiangying Dai, Xiaoru Dai, Xuan Huang
The paper introduces Reinforcement Learning Enhanced LLM Agents (RLEA), a multi‑agent framework that automates the modeling of complex Vehicle Routing Problems (VRPs). RLEA employs a lightweight neural Planner trained with Soft Q‑learning to coordinate LLM‑based agents, and incorporates an evolutionary memory module and retrieval‑augmented generation to leverage experience and external solver knowledge. Experiments on 48 VRP variants show that RLEA outperforms the prior state‑of‑the‑art method, achieving a 16.67% higher success rate and significantly reducing runtime errors.
By Yi Chen, Zikang Yu, Jiahai Wang, Jinbiao Chen, Jianpeng Zhou, Zizhen Zhang