CreativeBench: Benchmarking and Enhancing Machine Creativity via Self-Evolving Challenges
Zi-Han Wang, Lam Nguyen, Zhengyang Zhao, Mengyue Yang, Chengwei Qin, Yujiu Yang, Linyi Yang
[ArXiv] [Homepage] [Code]

• Introduces CreativeBench, a benchmark for machine creativity in code generation, covering both combinatorial and exploratory creativity with executable evaluation and an inference-time enhancement strategy, EvoRePE.

Alignment for Efficient Tool Calling of Large Language Models
Hongshen Xu*, Zihan Wang*, Zichen Zhu, Lei Pan, Xingyu Chen, Lu Chen, Kai Yu
[ArXiv]

• Proposes a multi-objective alignment framework combining probabilistic knowledge boundary estimation with dynamic decision making to reduce unnecessary tool calls while preserving performance.

Reducing Tool Hallucination via Reliability Alignment
Hongshen Xu, Su Zhu, Zihan Wang, Hang Zheng, Da Ma, Ruisheng Cao, Shuai Fan, Lu Chen, Kai Yu
[ArXiv]

• Defines and categorizes tool hallucinations (tool-selection vs tool-usage) and introduces reliability-oriented alignment to improve robust tool interaction and efficiency.

Memento: Fine-tuning LLM Agents without Fine-tuning LLMs
Huichi Zhou, Yihang Chen, Siyuan Guo, Xue Yan, Kin Hei Lee, Zihan Wang, Ka Yiu Lee, Guchun Zhang, Kun Shao, Linyi Yang†, Jun Wang†
[ArXiv] [Code]

• A memory-based continual improvement framework enabling agents to learn from experience without updating model weights.

Delusions of Large Language Models
Hongshen Xu, Zixv Yang, Zichen Zhu, Kunyao Lan, Zihan Wang, Mengyue Wu, Ziwei Ji, Lu Chen, Pascale Fung, Kai Yu
[ArXiv]

• Investigates high-confidence hallucination phenomena (“delusions”) and analyzes underlying causes and behaviors.