7 papers
Harnessing LLMs with Direction-Scope Control for Combinatorial Solver Refinement
Yili Cao, Anqi Li, Haoran Li +2
Harnesses are increasingly unlocking the potential of large language model (LLM) agents by making their actions more controllable. In automatic heuristic design (AHD) for combinato…
The Hidden Power of Scaling Factor in LoRA Optimization
Zicheng Zhang, Haoran Li, Jiaxing Wang +10
In Low-Rank Adaptation (LoRA), the scaling factor is often treated as a mere complement to the learning rate, yet its role in optimization remains poorly understood. In this pa…
On the Tension Between Optimality and Adversarial Robustness in Policy Optimization
Haoran Li, Jiayu Lv, Congying Han +5
Achieving optimality and adversarial robustness in deep reinforcement learning has long been regarded as conflicting goals. Nonetheless, recent theoretical insights presented in CA…
Parallel Sampling via Autospeculation
Nima Anari, Carlo Baronio, CJ Chen +4
We present parallel algorithms to accelerate sampling via counting in two settings: any-order autoregressive models and denoising diffusion models. An any-order autoregressive mode…
Purity Law for Generalizable Neural TSP Solvers
Wenzhao Liu, Haoran Li, Congying Han +3
Achieving generalization in neural approaches across different scales and distributions remains a significant challenge for the Traveling Salesman Problem~(TSP). A key obstacle is…
Adversarially-Robust Inference on Trees via Belief Propagation
Samuel B. Hopkins, Anqi Li
We introduce and study the problem of posterior inference on tree-structured graphical models in the presence of a malicious adversary who can corrupt some observed nodes. In the w…
General Method for Solving Four Types of SAT Problems
Anqi Li, Congying Han, Tiande Guo +2
Existing methods provide varying algorithms for different types of Boolean satisfiability problems (SAT), lacking a general solution framework. Accordingly, this study proposes a u…