Institution profile

Georgia Institute of Technology

Academic institutionnorthamerica · us
Official website
Research library2,814linked papers
Opportunities0open roles
Selected work

Representative Papers

A Survey on Large Language Model-Based Game Agents

Apr 02, 2024arXiv.org

This paper addresses the challenge of enabling human-like decision-making in game agents operating within complex environments. Methodologically, it proposes the first three-dimensional functional architecture—Memory–Reasoning–I/O—for LLM-driven game agents, systematically reviewing over 50 representative works across six game genres, including adventure, communication, and competitive games. The approach integrates multimodal perception, long-term memory, chain-of-thought reasoning, and game API integration to establish cross-genre unified evaluation dimensions. Key contributions include: (1) the first formalization of a functional architecture for LLM-based game agents; (2) the creation of an open-source, structured, and authoritative repository of relevant literature; and (3) an empirical analysis revealing critical performance bottlenecks and generalization limitations, thereby providing both a theoretical framework and a practical roadmap for AGI-oriented game agent research.

45 citations2 influentialRead paper

Imitating and Finetuning Model Predictive Control for Robust and Symmetric Quadrupedal Locomotion

Nov 01, 2023IEEE Robotics and Automation Letters

To address the challenge of achieving robust, symmetric, periodic, and energy-efficient gaits for quadrupedal robots on unstructured terrains—including uneven ground, slippery surfaces, and moving conveyor belts—this paper proposes an “Imitation–Fine-tuning” collaborative control framework (IFM). IFM innovatively employs a model predictive controller (MPC) combining differential dynamic programming with Raibert-inspired heuristics as an expert policy, which is first rendered learnable via behavioral cloning and subsequently refined safely using low-exploration-depth PPO or SAC reinforcement learning. Comprehensive simulation and hardware experiments demonstrate that IFM significantly improves gait stability and left–right symmetry, generates more periodic and energy-efficient locomotion patterns, and eliminates the need for intricate reward engineering. The framework achieves a favorable trade-off among safety, environmental adaptability, and deployment efficiency.

15 citationsRead paper

In-and-Out: Algorithmic Diffusion for Sampling Convex Bodies

May 02, 2024Neural Information Processing Systems

This work addresses efficient uniform sampling from high-dimensional convex bodies, providing strong convergence guarantees under Rényi divergence—including total variation (TV), Wasserstein-2 ($mathcal{W}_2$), Kullback–Leibler (KL), and $chi^2$ divergences. We propose a novel stochastic walk algorithm that models sampling through the lens of stochastic diffusion—a first in this context—and characterizes convergence rates via the functional isoperimetric constant of the target distribution, thereby departing from conventional polynomial mixing-time analysis. Theoretically, our algorithm achieves the optimal time complexity $O^*(n^2 R^2)$, where $n$ denotes dimensionality and $R$ the body’s diameter. Moreover, we derive unified, tight convergence bounds across the entire Rényi divergence family. To our knowledge, this is the first uniform sampling scheme attaining simultaneous optimality under multiple probability metrics.

12 citations2 influentialRead paper

Explicit Group Sparse Projection with Applications to Deep Learning and NMF

Dec 09, 2019Trans. Mach. Learn. Res.

This work addresses the challenge of explicitly controlling the average sparsity—measured by the Hoyer metric—in sparse projections of vector sets. We propose the first group-level explicit sparsity projection method: it directly specifies a target average sparsity level and jointly optimizes the sparsity patterns of all vectors, eliminating per-vector processing or reliance on implicit regularization parameters. Our approach generalizes the weighted ℓ₁ norm, enabling flexible sparsity modeling with linear-time computational complexity. The key innovation is the first formulation that imposes interpretable, tunable average sparsity constraints over an entire vector group. Experiments demonstrate substantial improvements in the accuracy–sparsity trade-off for ResNet50 pruning, and competitive reconstruction error and classification performance on CIFAR-10, ImageNet, and non-negative matrix factorization tasks.

10 citationsRead paper

H3Fusion: Helpful, Harmless, Honest Fusion of Aligned LLMs

Nov 26, 2024arXiv.org

This work addresses the limited performance gains and insufficient robustness in multi-model alignment fusion. We propose H³Fusion, a novel method built upon a Mixture-of-Experts (MoE) architecture, featuring an instruction-aware dynamic expert routing mechanism that fine-tunes only the feed-forward network (FFN) layers while keeping attention weights frozen. To the best of our knowledge, H³Fusion introduces the first triple-alignment fusion paradigm, jointly optimizing helpfulness, harmlessness, and honesty via a gated loss function and adaptive regularization—thereby co-regulating expert activation distributions and parameter drift. Evaluated on three core alignment benchmarks, H³Fusion achieves an average 11.37% improvement in alignment performance and demonstrates 13.77% higher robustness than the current state-of-the-art ensemble methods, significantly enhancing model outputs’ helpfulness, harmlessness, and honesty.

7 citations2 influentialRead paper
Recent publications

Latest Papers

Online Allocation using Few Samples

Sep 15, 2026

研究在线分配问题，通过少量样本在对抗顺序下实现资源的有效分配或负载均衡。提出一种通用框架将随机顺序模型算法转换为适用于更复杂采样模型的算法。

0 citationsRead paper