arXiv subjects

Chaoxu Mu

Publications and source records attributed to Chaoxu Mu.

12 recordsLinked to original sources

arXiv2026

UP-NRPA: User Portrait based Nested Rollout Policy Adaptation for Planning with Large Language Models in Goal-oriented Dialogue Systems

To address the challenge that current dialogue policy planning methods struggle to dynamically adapt to diverse user characteristics, this paper proposes a User Portrait based Nested Rollout Policy Adaptation (UP-NRPA) online framework with Large Language Models. In contrast to conventional approaches dependent on model training and require offline reinforcement learning policy models for user groups, UP-NRPA enables dynamic customization of dialogue strategies through an adaptive mechanism. This is achieved by leveraging real-time user feedback alongside personality, preferences, and objectives mapped from the current user portrait, thereby adapting to user characteristics without offline reinforcement learning. In collaborative and non-collaborative dialogue benchmarks, UP-NRPA demonstrated considerable benefits, achieving an impressive 100% success rate in multiple dialogue tasks. Particularly in negotiation tasks, the sale-to-list ratio (SL) increased by 56.41%. This demonstrates that UP-NRPA can adapt to diverse user needs without requiring a training mechanism, enabling the dialogue system to adapt to user characteristics.

cs.AI↗

arXiv2026

Spatially Generalizable Mobile Manipulation via Adaptive Experience Selection and Dynamic Imagination

Mobile Manipulation (MM) involves long-horizon decision-making over multi-stage compositions of heterogeneous skills, such as navigation and picking up objects. Despite recent progress, existing MM methods still face two key limitations: (i) low sample efficiency, due to ineffective use of redundant data generated during long-term MM interactions; and (ii) poor spatial generalization, as policies trained on specific tasks struggle to transfer to new spatial layouts without additional training. In this paper, we address these challenges through Adaptive Experience Selection (AES) and model-based dynamic imagination. In particular, AES makes MM agents pay more attention to critical experience fragments in long trajectories that affect task success, improving skill chain learning and mitigating skill forgetting. Based on AES, a Recurrent State-Space Model (RSSM) is introduced for Model-Predictive Forward Planning (MPFP) by capturing the coupled dynamics between the mobile base and the manipulator and imagining the dynamics of future manipulations. RSSM-based MPFP can reinforce MM skill learning on the current task while enabling effective generalization to new spatial layouts. Comparative studies across different experimental configurations demonstrate that our method significantly outperforms existing MM policies. Real-world experiments further validate the feasibility and practicality of our method.

cs.RO↗

arXiv2026

DisCo-FLoc: Semantic-Free Floorplan Localization via $SE(2)$-Aware Contrastive Disambiguation

Visual Floorplan Localization (FLoc) struggles with severe structural aliasing caused by repetitive minimalist layouts. This occurs because physically distant poses share highly similar visual-geometric features, which degrades spatial separability and angular discriminability. While existing methods attempt to mitigate these ambiguities by relying on costly semantic annotations, the resulting performance gains remain inherently limited. To address the above issues, we propose DisCo-FLoc, a semantic-free method for visual-geometric Contrastive Disambiguation. First, we introduce a depth-aware Ray Regression Predictor (RRP) that serves as a dense-to-ray geometric projector. By explicitly suppressing visual clutter along the vertical dimension, RRP projects monocular RGB images into 2D ray primitives, which are matched with floorplans to produce geometry-aware FLoc candidates. Second, to resolve the remaining ambiguity among these candidates, we propose a spatially perturbed contrastive objective to align RGB images with local floorplan structures and formulate a visual-geometric compatibility function. In particular, we meticulously construct positive and negative samples at both positional and directional levels through $SE(2)$ pose perturbations for contrastive learning, effectively achieving pose smoothness, spatial separability, and angular discriminability. The compatibility function enables DisCo-FLoc to disambiguate FLoc by using richer visual context beyond pure geometric layouts, without requiring any semantic annotations. Extensive experiments on two challenging visual FLoc benchmarks demonstrate that DisCo-FLoc significantly outperforms state-of-the-art semantic-based methods, especially narrowing the performance gap between positional and directional FLoc accuracy.

cs.RO↗

arXiv2025

CogMCTS: A Novel Cognitive-Guided Monte Carlo Tree Search Framework for Iterative Heuristic Evolution with Large Language Models

Automatic Heuristic Design (AHD) is an effective framework for solving complex optimization problems. The development of large language models (LLMs) enables the automated generation of heuristics. Existing LLM-based evolutionary methods rely on population strategies and are prone to local optima. Integrating LLMs with Monte Carlo Tree Search (MCTS) improves the trade-off between exploration and exploitation, but multi-round cognitive integration remains limited and search diversity is constrained. To overcome these limitations, this paper proposes a novel cognitive-guided MCTS framework (CogMCTS). CogMCTS tightly integrates the cognitive guidance mechanism of LLMs with MCTS to achieve efficient automated heuristic optimization. The framework employs multi-round cognitive feedback to incorporate historical experience, node information, and negative outcomes, dynamically improving heuristic generation. Dual-track node expansion combined with elite heuristic management balances the exploration of diverse heuristics and the exploitation of high-quality experience. In addition, strategic mutation modifies the heuristic forms and parameters to further enhance the diversity of the solution and the overall optimization performance. The experimental results indicate that CogMCTS outperforms existing LLM-based AHD methods in stability, efficiency, and solution quality.

cs.AI↗

arXiv2025

A General Highly Accurate Online Planning Method Integrating Large Language Models into Nested Rollout Policy Adaptation for Dialogue Tasks

In goal-oriented dialogue tasks, the main challenge is to steer the interaction towards a given goal within a limited number of turns. Existing approaches either rely on elaborate prompt engineering, whose effectiveness is heavily dependent on human experience, or integrate policy networks and pre-trained policy models, which are usually difficult to adapt to new dialogue scenarios and costly to train. Therefore, in this paper, we present Nested Rollout Policy Adaptation for Goal-oriented Dialogue (NRPA-GD), a novel dialogue policy planning method that completely avoids specific model training by utilizing a Large Language Model (LLM) to simulate behaviors of user and system at the same time. Specifically, NRPA-GD constructs a complete evaluation mechanism for dialogue trajectories and employs an optimization framework of nested Monte Carlo simulation and policy self-adaptation to dynamically adjust policies during the dialogue process. The experimental results on four typical goal-oriented dialogue datasets show that NRPA-GD outperforms both existing prompt engineering and specifically pre-trained model-based methods. Impressively, NRPA-GD surpasses ChatGPT and pre-trained policy models with only a 0.6-billion-parameter LLM. The proposed approach further demonstrates the advantages and novelty of employing planning methods on LLMs to solve practical planning tasks.

cs.CL↗

arXiv2025

MMWOZ: Building Multimodal Agent for Task-oriented Dialogue

Task-oriented dialogue systems have garnered significant attention due to their conversational ability to accomplish goals, such as booking airline tickets for users. Traditionally, task-oriented dialogue systems are conceptualized as intelligent agents that interact with users using natural language and have access to customized back-end APIs. However, in real-world scenarios, the widespread presence of front-end Graphical User Interfaces (GUIs) and the absence of customized back-end APIs create a significant gap for traditional task-oriented dialogue systems in practical applications. In this paper, to bridge the gap, we collect MMWOZ, a new multimodal dialogue dataset that is extended from MultiWOZ 2.3 dataset. Specifically, we begin by developing a web-style GUI to serve as the front-end. Next, we devise an automated script to convert the dialogue states and system actions from the original dataset into operation instructions for the GUI. Lastly, we collect snapshots of the web pages along with their corresponding operation instructions. In addition, we propose a novel multimodal model called MATE (Multimodal Agent for Task-oriEnted dialogue) as the baseline model for the MMWOZ dataset. Furthermore, we conduct comprehensive experimental analysis using MATE to investigate the construction of a practical multimodal agent for task-oriented dialogue.

cs.CL↗

arXiv2025

Planning of Heuristics: Strategic Planning on Large Language Models with Monte Carlo Tree Search for Automating Heuristic Optimization

Heuristics have achieved great success in solving combinatorial optimization problems~(COPs). However, heuristics designed by humans require too much domain knowledge and testing time. Since Large Language Models~(LLMs) possess strong capabilities to understand and generate content with a knowledge base that covers various domains, they offer potential ways to automatically optimize heuristics. To this end, we propose Planning of Heuristics~(PoH), an optimization method that integrates LLM self-reflection with Monte Carlo Tree Search, a well-known planning algorithm. PoH iteratively refines generated heuristics by evaluating their performance and providing improvement suggestions. Our method enables to iteratively evaluate the generated heuristics~(states) and improve them based on the improvement suggestions~(actions) and evaluation results~(rewards), by effectively simulating future states to search for paths with higher rewards. In this paper, we apply PoH to solve the Traveling Salesman Problem and the Flow Shop Scheduling Problem. The experimental results show that PoH outperforms hand-crafted heuristics and other Automatic Heuristic Design methods based on LLMs, and achieves the state-of-the-art performance in automating heuristic optimization with LLMs to solve tested COPs, especially with large sizes.

cs.AI↗

arXiv2025

TSS GAZ PTP: Towards Improving Gumbel AlphaZero with Two-stage Self-play for Multi-constrained Electric Vehicle Routing Problems

Recently, Gumbel AlphaZero~(GAZ) was proposed to solve classic combinatorial optimization problems such as TSP and JSSP by creating a carefully designed competition model~(consisting of a learning player and a competitor player), which leverages the idea of self-play. However, if the competitor is too strong or too weak, the effectiveness of self-play training can be reduced, particularly in complex CO problems. To address this problem, we further propose a two-stage self-play strategy to improve the GAZ method~(named TSS GAZ PTP). In the first stage, the learning player uses the enhanced policy network based on the Gumbel Monte Carlo Tree Search~(MCTS), and the competitor uses the historical best trained policy network~(acts as a greedy player). In the second stage, we employ Gumbel MCTS for both players, which makes the competition fiercer so that both players can continuously learn smarter trajectories. We first investigate the performance of our proposed TSS GAZ PTP method on TSP since it is also used as a test problem by the original GAZ. The results show the superior performance of TSS GAZ PTP. Then we extend TSS GAZ PTP to deal with multi-constrained Electric Vehicle Routing Problems~(EVRP), which is a recently well-known real application research topic and remains challenging as a complex CO problem. Impressively, the experimental results show that the TSS GAZ PTP outperforms the state-of-the-art Deep Reinforcement Learning methods in all types of instances tested and outperforms the optimization solver in tested large-scale instances, indicating the importance and promising of employing more dynamic self-play strategies for complex CO problems.

eess.SY↗

arXiv2025

One for All: A General Framework of LLMs-based Multi-Criteria Decision Making on Human Expert Level

Multi-Criteria Decision Making~(MCDM) is widely applied in various fields, using quantitative and qualitative analyses of multiple levels and attributes to support decision makers in making scientific and rational decisions in complex scenarios. However, traditional MCDM methods face bottlenecks in high-dimensional problems. Given the fact that Large Language Models~(LLMs) achieve impressive performance in various complex tasks, but limited work evaluates LLMs in specific MCDM problems with the help of human domain experts, we further explore the capability of LLMs by proposing an LLM-based evaluation framework to automatically deal with general complex MCDM problems. Within the framework, we assess the performance of various typical open-source models, as well as commercial models such as Claude and ChatGPT, on 3 important applications, these models can only achieve around 60\% accuracy rate compared to the evaluation ground truth. Upon incorporation of Chain-of-Thought or few-shot prompting, the accuracy rates rise to around 70\%, and highly depend on the model. In order to further improve the performance, a LoRA-based fine-tuning technique is employed. The experimental results show that the accuracy rates for different applications improve significantly to around 95\%, and the performance difference is trivial between different models, indicating that LoRA-based fine-tuned LLMs exhibit significant and stable advantages in addressing MCDM tasks and can provide human-expert-level solutions to a wide range of MCDM challenges.

cs.AI↗

arXiv2025

Reducing Action Space for Deep Reinforcement Learning via Causal Effect Estimation

Intelligent decision-making within large and redundant action spaces remains challenging in deep reinforcement learning. Considering similar but ineffective actions at each step can lead to repetitive and unproductive trials. Existing methods attempt to improve agent exploration by reducing or penalizing redundant actions, yet they fail to provide quantitative and reliable evidence to determine redundancy. In this paper, we propose a method to improve exploration efficiency by estimating the causal effects of actions. Unlike prior methods, our approach offers quantitative results regarding the causality of actions for one-step transitions. We first pre-train an inverse dynamics model to serve as prior knowledge of the environment. Subsequently, we classify actions across the entire action space at each time step and estimate the causal effect of each action to suppress redundant actions during exploration. We provide a theoretical analysis to demonstrate the effectiveness of our method and present empirical results from simulations in environments with redundant actions to evaluate its performance. Our implementation is available at https://github.com/agi-brain/cee.git.

cs.LG↗

arXiv2024

Soft Masked Transformer for Point Cloud Processing with Skip Attention-Based Upsampling

Point cloud processing methods leverage local and global point features %at the feature level to cater to downstream tasks, yet they often overlook the task-level context inherent in point clouds during the encoding stage. We argue that integrating task-level information into the encoding stage significantly enhances performance. To that end, we propose SMTransformer which incorporates task-level information into a vector-based transformer by utilizing a soft mask generated from task-level queries and keys to learn the attention weights. Additionally, to facilitate effective communication between features from the encoding and decoding layers in high-level tasks such as segmentation, we introduce a skip-attention-based up-sampling block. This block dynamically fuses features from various resolution points across the encoding and decoding layers. To mitigate the increase in network parameters and training time resulting from the complexity of the aforementioned blocks, we propose a novel shared position encoding strategy. This strategy allows various transformer blocks to share the same position information over the same resolution points, thereby reducing network parameters and training time without compromising accuracy.Experimental comparisons with existing methods on multiple datasets demonstrate the efficacy of SMTransformer and skip-attention-based up-sampling for point cloud processing tasks, including semantic segmentation and classification. In particular, we achieve state-of-the-art semantic segmentation results of 73.4% mIoU on S3DIS Area 5 and 62.4% mIoU on SWAN dataset

cs.CV↗

arXiv2023

XuanCe: A Comprehensive and Unified Deep Reinforcement Learning Library

In this paper, we present XuanCe, a comprehensive and unified deep reinforcement learning (DRL) library designed to be compatible with PyTorch, TensorFlow, and MindSpore. XuanCe offers a wide range of functionalities, including over 40 classical DRL and multi-agent DRL algorithms, with the flexibility to easily incorporate new algorithms and environments. It is a versatile DRL library that supports CPU, GPU, and Ascend, and can be executed on various operating systems such as Ubuntu, Windows, MacOS, and EulerOS. Extensive benchmarks conducted on popular environments including MuJoCo, Atari, and StarCraftII multi-agent challenge demonstrate the library's impressive performance. XuanCe is open-source and can be accessed at https://github.com/agi-brain/xuance.git.

cs.LG↗