Publications
Conference and journal papers, in reverse chronological order.
2026
- ICLR 2026
Towards Self-Robust LLMs: Intrinsic Prompt Noise Resistance via CoIPOIn The Fourteenth International Conference on Learning Representations, 2026 - ICML 2026
cMoLLM at Scale: Horizontal Scaling Laws for Mixture-of-LLMsIn Forty-Third International Conference on Machine Learning, 2026 - ICML 2026
Position: Multi-Agent Systems Should Prioritize Concurrency ControlIn Forty-third International Conference on Machine Learning Position Paper Track, 2026 - CVPR 2026
Breaking the Illusion: When Positive Meets Negative in Multimodal DecodingIn The IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2026 - Tech Report
Higher Satisfaction, Lower Cost: A Technical Report on How LLMs Revolutionize Meituan’s Intelligent Interaction SystemsIn Meituan Technical Report, 2026 - ICML 2026 Workshop
Toward Trustworthy LLM Router Ecosystems: Incentive-Compatible Cryptographic MitigationsIn Trustworthy AI for Good (AI4GOOD) Workshop @ ICML 2026, 2026 - arXiv
V-tableR1: Process-Supervised Multimodal Table Reasoning with Critic-Guided Policy Optimization2026 - arXiv
HDSL: A Hierarchical Domain-Specific Language for Structured 3D Indoor Scene Generation and Localized Editing with LLM Agents2026
2025
- IEEE SMC 2025
Can LLMs Write Fast System-Aware Numerical Computation Code?In 2025 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2025 - PRICAI 2025
Leveraging Runtime Information for LLM QuantizationIn PRICAI 2025: Trends in Artificial Intelligence, 2025 - ICML 2025 Workshop
INTEGRALBENCH: Benchmarking LLMs with Definite Integral ProblemsIn 2nd AI for Math Workshop @ ICML, 2025 - arXiv
Towards Provable (In)Secure Model Weight Release Schemes2025