11 citations · 24 across the 33 of their papers we have counts for
37 papers
Verifier-Harness Co-Evolution for Training-Free Self-Improvement on Open-Ended Tasks
Tianyi Xu, Yaolun Zhang, Yiran Wu +6
Training-free self-improvement lets an agent use feedback to revise the response it is about to return and to update the harness state that shapes how it handles later tasks. On op…
EvalResearchBench: Can AI Agents Build Their Own Feedback for Recursive Self-Improvement?
Yaolun Zhang, Tianyi Xu, Yujie Zhao +3
Recursive self-improvement (RSI) relies on evaluation feedback to assess progress and guide further research, yet repeatedly running complex benchmarks is costly and slows iteratio…
Pin Once, Swap Light: Subspace-Aligned Centroid-Residual Training for Efficient Ultra-LoRA Serving
Xiang Li, Pengcheng Wang, Huazheng Wang +1
Modern multi-tenant Low-Rank Adapters (LoRAs) serving systems concurrently host tens to hundreds of LoRA adapters. Though powerful, this introduces a critical system dilemma betwee…
From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement
Qinsi Wang, Jing Shi, Huazheng Wang +8
Reinforcement Learning with Verifiable Rewards (RLVR) has driven recent progress in reasoning-oriented large language models (LLMs) by enabling large-scale optimization. However, i…
Online KL-Regularized Reinforcement Learning with Function Approximation under Misspecification
Haoyang Hong, Zichen Wang, Quanquan Gu +1
We study KL-regularized contextual bandits and episodic reinforcement learning (RL) under general function approximation with model misspecification. Existing guarantees rely on re…
Who&When Pro: Can LLMs Really Attribute Failures in AI Agents?
Jiale Liu, Huajun Xi, Shaokun Zhang +6
Automated failure attribution uses LLMs to identify where and why agentic systems fail. As agents become more capable, their failures become subtler, making automated attribution i…
EvoPool: Evolutionary Programmatic Annotation for Label-Efficient Specialized Supervision
Tianyi Xu, Yaolun Zhang, Xuan Ouyang +1
Large language models excel at general tasks but underperform smaller supervised models in specialized, high-stakes domains where training labels are costly. We address this regime…
When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs
Yifan Zeng, Yiran Wu, Yaolun Zhang +4
Multi-agent LLM workflows route inference through specialized roles to lift end-task accuracy, but jointly training those roles with reinforcement learning is unstable in ways that…
MetaAgent-X : Breaking the Ceiling of Automatic Multi-Agent Systems via End-to-End Reinforcement Learning
Yaolun Zhang, Yujie Zhao, Nan Wang +6
Automatic multi-agent systems aim to instantiate agent workflows without relying on manually designed or fixed orchestration. However, existing automatic MAS approaches remain only…
EVOCHAMBER: Test-Time Co-evolution of Multi-Agent System at Individual, Team, and Population Scales
Yaolun Zhang, Tianyi Xu, Shengyu Dai +3
We argue that multi-agent test-time evolution is not single-agent evolution replicated N times. A single-agent learner can only evolve its own context and memory. A multi-agent sys…
Beyond Static Bias: Adaptive Multi-Fidelity Bandits with Improving Proxies
Muyun Lu, Haoyang Hong, Huazheng Wang +1
As an extension of the classical multi-armed bandit problem, multi-fidelity multi-armed bandits (MF-MAB) enable individual arms to be evaluated using diverse feedback sources that…
Speculative Pipeline Decoding: Higher-Accuracy Drafting with Hidden Latency via Pipeline Parallelism
Yijiong Yu, Huazheng Wang, Shuai Yuan +2
Speculative Decoding (SD) accelerates low-concurrency LLM inference with a draft-then-verify paradigm. Mainstream methods, however, rely on multi-token prediction, which incurs com…
Towards Affordable Energy: A Gymnasium Environment for Electric Utility Demand-Response Programs
Jose E. Aguilar Escamilla, Lingdong Zhou, Xiangqi Zhu +1
Extreme weather and volatile wholesale electricity markets expose residential consumers to catastrophic financial risks, yet demand response at the distribution level remains an un…
When Can You Poison Rewards? A Tight Characterization of Reward Poisoning in Linear MDPs
Jose Efraim Aguilar Escamilla, Haoyang Hong, Jiawei Li +4
We study reward poisoning attacks in reinforcement learning (RL), where an adversary manipulates rewards within constrained budgets to force the target RL agent to adopt a policy t…
Personalizing Mathematical Game-based Learning for Children: A Preliminary Study
Jie Gao, Adam K. Dubé
Game-based learning (GBL) is widely adopted in mathematics education. It enhances learners' engagement and critical thinking throughout the mathematics learning process. However, e…
Live-Evo: Online Evolution of Agentic Memory from Continuous Feedback
Yaolun Zhang, Yiran Wu, Yijiong Yu +2
Large language model (LLM) agents are increasingly equipped with memory, which are stored experience and reusable guidance that can improve task-solving performance. Recent \emph{s…
SWAA: Sliding Window Attention Adaptation for Efficient and Quality Preserving Long Context Processing
Yijiong Yu, Jiale Liu, Qingyun Wu +2
The quadratic complexity of self attention in Transformer based LLMs renders long context inference prohibitively expensive. While Sliding Window Attention (SWA), the simplest spar…
Design-Based Bandits Under Network Interference: Trade-Off Between Regret and Statistical Inference
Zichen Wang, Haoyang Hong, Chuanhao Li +3
In multi-armed bandits with network interference (MABNI), the action taken by one node can influence the rewards of others, creating complex interdependence. While existing researc…
The Ranking Blind Spot: Decision Hijacking in LLM-based Text Ranking
Yaoyao Qian, Yifan Zeng, Yuchao Jiang +2
Large Language Models (LLMs) have demonstrated strong performance in information retrieval tasks like passage ranking. Our research examines how instruction-following capabilities…
A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence
Huan-ang Gao, Jiayi Geng, Wenyue Hua +24
Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse tasks but remain fundamentally static, unable to adapt their internal parameters to novel task…
SimpleDoc: Multi-Modal Document Understanding with Dual-Cue Page Retrieval and Iterative Refinement
Chelsi Jain, Yiran Wu, Yifan Zeng +5
Document Visual Question Answering (DocVQA) is a practical yet challenging task, which is to ask questions based on documents while referring to multiple pages and different modali…
Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems
Shaokun Zhang, Ming Yin, Jieyu Zhang +8
Failure attribution in LLM multi-agent systems-identifying the agent and step responsible for task failures-provides crucial clues for systems debugging but remains underexplored a…
Provably Efficient Algorithm for Best Scoring Rule Identification in Online Principal-Agent Information Acquisition
Zichen Wang, Chuanhao Li, Huazheng Wang
We investigate the problem of identifying the optimal scoring rule within the principal-agent framework for online information acquisition problem. We focus on the principal's pers…
Divide, Optimize, Merge: Fine-Grained LLM Agent Optimization at Scale
Jiale Liu, Yifan Zeng, Shaokun Zhang +5
LLM-based optimization has shown remarkable potential in enhancing agentic systems. However, the conventional approach of prompting LLM optimizer with the whole training trajectori…