14 papers
RIPPLE: Recovery Internalization through Process-Level Policy Learning and Effect-Aware Credit for Long-Horizon Agentic Coding
Jingxuan Xu, Zihang Xu, Yuechen Luo +9
Long-horizon coding agents can observe a failure and generate a plausible critique yet continue along the same incorrect trajectory. We call this the Critique-to-Action Handoff Gap…
OPRD: On-Policy Reminder Distillation for Long-Horizon Coding Agents
Yuechen Luo, Zhiyi Lai, Miao Ziqi +9
Reliable repository-level coding requires agents not only to generate correct code but also to preserve constraints and respond to tool feedback across long interactions. However,…
SWE-Nexus: Automated Composition of Software Engineering Tasks from Test Execution
Songwei Yu, Ziqi Miao, Jingxuan Xu +10
Historical code changes provide convenient task boundaries, but do not identify which production functions selected tests exercise or support composing validated behaviors beyond a…
Defense-as-Skill: Evolving Runtime Guard Skill for Skill-Augmented Agents
Xiaofang Yang, Ziqi Miao, Chen Qian +3
Skill-augmented agents load reusable skills as persistent runtime context, improving task performance but also giving malicious skills a durable channel for steering future actions…
ReCap: Retrieval-Guided Refinement for Image Captioning Enhancement via Reinforcement Learning
Haonan Jia, Shichao Dong, Zenghui Sun +7
Reinforcement Learning (RL) has demonstrated significant gains in image captioning, yet it is still limited in encouraging Large Vision-Language Models (LVLMs) to explore novel rea…
Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations
Xinshun Feng, Ziqi Miao, Lijun Li +1
Large language model (LLM) agents are increasingly deployed in scientific research, where reliability is critical and the underlying knowledge is densely interconnected. In such se…
Modeling Complex Behaviors: Multi-Personality Composition and Dynamic Switching in Vision-Language Models
Peiqi Jia, Haonan Jia, Ziqi Miao +3
With the widespread deployment of Multimodal Large Language Models (MLLMs) in social interaction, understanding and controlling their behavior under complex personality conditions…
SkillPyramid: A Hierarchical Skill Consolidation Framework for Self-Evolving Agents
Yuan Xiong, Ziqi Miao, Qian Chen +6
Recent AI agents can flexibly invoke skills to solve complex tasks, but their long-term improvement is fundamentally constrained by a lack of systematic skill construction, accumul…
Seeing with You: Perception-Reasoning Coevolution for Multimodal Reasoning
Ziqi Miao, Haonan Jia, Lijun Li +4
Reinforcement learning with verifiable rewards (RLVR) has substantially enhanced the reasoning capabilities of multimodal large language models (MLLMs). However, existing RLVR appr…
Stable Adaptive Thinking via Advantage Shaping and Length-Aware Gradient Regulation
Zihang Xu, Haozhi Xie, Ziqi Miao +3
Large reasoning models (LRMs) achieve strong performance through extended reasoning traces, but they often exhibit overthinking behavior for low-complexity queries. Existing effort…
DeepSight: An All-in-One LM Safety Toolkit
Bo Zhang, Jiaxuan Guo, Lijun Li +17
As the development of Large Models (LMs) progresses rapidly, their safety is also a priority. In current Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) s…
TabSieve: Explicit In-Table Evidence Selection for Tabular Prediction
Yongyao Wang, Ziqi Miao, Lu Yang +4
Tabular prediction can benefit from in-table rows as few-shot evidence, yet existing tabular models typically perform instance-wise inference and LLM-based prompting is often britt…
Contextual Image Attack: How Visual Context Exposes Multimodal Safety Vulnerabilities
Yuan Xiong, Ziqi Miao, Lijun Li +3
While Multimodal Large Language Models (MLLMs) show remarkable capabilities, their safety alignments are susceptible to jailbreak attacks. Existing attack methods typically focus o…
Response Attack: Exploiting Contextual Priming to Jailbreak Large Language Models
Ziqi Miao, Lijun Li, Yuan Xiong +3
Contextual priming, where earlier stimuli covertly bias later judgments, offers an unexplored attack surface for large language models (LLMs). We uncover a contextual priming vulne…