Evaluate / User preferences
RouteJudge
Evaluate LLM routers through the lens of different user preferences, rather than a single shared priority.
About me
I am a master's student at the School of Artificial Intelligence, Nanjing University, and a member of the LAMDA Group, supervised by Prof. Han-Jia Ye.
My research focuses on reliable and efficient model reuse. I study how learning systems adapt continuously in dynamic environments, and how they select and reuse suitable capabilities from existing models. My recent work spans LLM routing, continual learning, test-time adaptation, and lifelong agents under resource constraints.
Three entry points into my routing research: an evaluation platform, an open-source toolbox, and a multimodal benchmark.
Evaluate / User preferences
Evaluate LLM routers through the lens of different user preferences, rather than a single shared priority.
Experiment / Open source
A toolbox for standardized LLM routing research and budget-aware evaluation, bringing routing experiments into a common framework.
Benchmark / Multimodal
A benchmark for studying how to route among multimodal LLMs, with a public dataset and code for further research.
I study reliable and efficient model reuse under changing tasks and limited resources. My earlier work helps individual models adapt through continual learning and test-time adaptation; my recent work explores how LLM routing can select, reuse, and reliably evaluate capabilities across multiple models.
Continual learning and test-time adaptation for individual models under changing tasks and environments.
Capability selection, routing supervision, and reliable evaluation of multiple models under resource constraints and user preferences.
NeurIPS'26
NeurIPS 2026
Examining why LLM routers systematically favor stronger, more expensive models, and how decision-aware routing can reduce unnecessary over-selection while preserving performance.
EMNLP'26
ECCV'26
CVPR'26
ICLR'26
CVPR'25
NeurIPS-W'26
NeurIPS 2026 Workshop on Towards Test-Time Continual Learning Agents
TokenWorth learns query-dependent memory worth from sparse execution feedback, guiding experience retrieval and bounded memory management for lifelong agents under constrained generation, interaction, and memory budgets.
ICMLW'26
Pluralistic Alignment Workshop at ICML 2026
Preference-based evaluation of LLM routers that accounts for diverse user priorities.
FCS'26
arXiv'26
arXiv'26
Started research on learnware and LLM routing.
School of Artificial Intelligence, LAMDA Group. Supervised by Prof. Han-Jia Ye; admitted without entrance examination.
Intelligent Finance.