Han Liu

A language model for statements of software code

2017

Building language models for source code enables a large set of improvements on traditional software engineering tasks. One promising application is automatic code completion. State-of-the-art techniques capture code regularities at token level with lexical information. …

A Heuristic T-S Fuzzy Model for the Pumped-Storage Generator-Motor Using Variable-Length Tree-Seed Algorithm-Based Competitive Agglomeration

2018 · Energies

With the fast development of artificial intelligence techniques, data-driven modeling approaches are becoming hotspots in both academic research and engineering practice. This paper proposes a novel data-driven T-S fuzzy model to precisely describe the complicated …

Marginal Policy Gradients for Complex Control.

2018 · arXiv (Cornell University)

Many complex domains, such as robotics control and real-time strategy (RTS) games, require an agent to learn a continuous control. In the former, an agent learns a policy over $\mathbb{R}^d$ and in the latter, over …

Joint measurement of time-frequency entanglement via sum frequency feedback

2019 · arXiv (Cornell University)

We propose, analyze and evaluate a technique for the joint measurement of time-frequency entanglement between two photons. In particular, we show that the frequency sum and time difference of two photons could be simultaneously measured …

Neural Polysynthetic Language Modelling.

2020

Many techniques in modern computational linguistics and natural language processing (NLP) make the assumption that approaches that work well on English and other widely used European (and sometimes Asian) languages are “language agnostic” – that …

Aligning Offline Metrics and Human Judgments of Value for Code Generation Models

2022 · arXiv (Cornell University)

Large language models have demonstrated great potential to assist programmers in generating code. For such human-AI pair programming scenarios, we empirically demonstrate that while generated code is most often evaluated in terms of their functional …

Section Location for Ground Faults Based on Primary-secondary Fusion and Current-heat Conversion

2024

A method based on the fusion of primary and secondary, as well as the conversion of current and heat, is introduced to solve the issue of locating the ground fault section. Specifically, a line terminal …

Improving Accuracy and Calibration via Differentiated Deep Mutual Learning

2025

Deep Neural Networks (DNNs) have achieved remarkable success in a variety of tasks, particularly in terms of prediction accuracy. However, in real-world scenarios, especially in safety-critical applications, accuracy alone is insufficient; reliable uncertainty estimates are …

AdvEvo-MARL: Shaping Internalized Safety through Adversarial Co-Evolution in Multi-Agent Reinforcement Learning

2025 · arXiv (Cornell University)

LLM-based multi-agent systems excel at planning, tool use, and role coordination, but their openness and interaction complexity also expose them to jailbreak, prompt-injection, and adversarial collaboration. Existing defenses fall into two lines: (i) self-verification that …

Label Efficient Semi-Supervised Learning via Graph Filtering

2019

Graph-based methods have been demonstrated as one of the most effective approaches for semi-supervised learning, as they can exploit the connectivity patterns between labeled and unlabeled data samples to improve learning performance. However, existing graph-based …