Huanyu Li

Dongyu Ru, Lin Qiu, Xiangkun Hu et al. · amazon-science

Despite Retrieval-Augmented Generation (RAG) showing promising capability in leveraging external knowledge, a comprehensive evaluation of RAG systems is still challenging due to the modular nature of RAG, evaluation of long-form responses and reliability of measurements. In this paper, we propose a fine-grained evaluation framework, RAGChecker, that incorporates a suite of diagnostic metrics for both the retrieval and generation modules. Meta evaluation verifies that RAGChecker has significantly better correlations with human judgments than other evaluation metrics. Using RAGChecker, we evaluate 8 RAG systems and conduct an in-depth analysis of their performance, revealing insightful patterns and trade-offs in the design choices of RAG architectures. The metrics of RAGChecker can guide researchers and practitioners in developing more effective RAG systems. This work has been open sourced at https://github.com/amazon-science/RAGChecker.

Huanyu Li, Els de Vleeschauwer, Robin Keskisärkkä et al.

Increasing the circularity of resource use in our society has been recognized as a path to sustainability, i.e., transitioning into a more circular economy. There are many different circular strategies to do so, such as reusing products and components, refurbishing and remanufacturing used products, or recycling left-over or used materials. To enable these strategies, it is necessary to share information at the infrastructure level and to communicate between industry sectors along the product life cycle. Enabling semantic interoperability in this information sharing and communication is therefore a key to increasing circularity. However, knowledge representation for the circular economy (CE) domain, which involves many relevant industry sectors related to product life cycles, remains challenging. To bridge this gap, we developed the Circular Economy Ontology Network (CEON) within the Onto-DESIDE project. This ontology network aims to fill gaps in CE by defining cross-sectorial concepts and to enable semantics-aware data documentation. We demonstrate CEON through cross-industry data documentation scenarios spanning construction, electronics, and textile sectors.

Han Xue, Yutong Li, Wenqiang Xu et al.

This paper explores the development of UniFolding, a sample-efficient, scalable, and generalizable robotic system for unfolding and folding various garments. UniFolding employs the proposed UFONet neural network to integrate unfolding and folding decisions into a single policy model that is adaptable to different garment types and states. The design of UniFolding is based on a garment's partial point cloud, which aids in generalization and reduces sensitivity to variations in texture and shape. The training pipeline prioritizes low-cost, sample-efficient data collection. Training data is collected via a human-centric process with offline and online stages. The offline stage involves human unfolding and folding actions via Virtual Reality, while the online stage utilizes human-in-the-loop learning to fine-tune the model in a real-world setting. The system is tested on two garment types: long-sleeve and short-sleeve shirts. Performance is evaluated on 20 shirts with significant variations in textures, shapes, and materials. More experiments and videos can be found in the supplementary materials and on the website: https://unifolding.robotflow.ai

Yuankai Zhou, Huanyu Li

Non-core drilling has gradually become the primary exploration method in geological exploration engineering, and well logging curves have increasingly gained importance as the main carriers of geological information. However, factors such as geological environment, logging equipment, borehole quality, and unexpected events can all impact the quality of well logging curves. Previous methods of re-logging or manual corrections have been associated with high costs and low efficiency. This paper proposes a machine learning method that utilizes existing data to predict missing data, and its effectiveness and feasibility have been validated through field experiments. The proposed method builds on the traditional Long Short-Term Memory (LSTM) neural network by incorporating a self-attention mechanism to analyze the sequential dependencies of the data. It selects the dominant computational results in the LSTM, reducing the computational complexity from O(n^2) to O(nlogn) and improving model efficiency. Experimental results demonstrate that the proposed method achieves higher accuracy compared to traditional curve synthesis methods based on Fully Connected Neural Networks (FCNN) and vanilla LSTM. This accurate, efficient, and cost-effective prediction method holds a practical value in engineering applications.

Huanyu Li, Dewei Wang, Xinmiao Wang et al.

Humanoid robots often need to balance competing objectives, such as maximizing speed while minimizing energy consumption. While current reinforcement learning (RL) methods can master complex skills like fall recovery and perceptive locomotion, they are constrained by fixed weighting strategies that produce a single suboptimal policy, rather than providing a diverse set of solutions for sophisticated multi-objective control. In this paper, we propose a novel framework leveraging Multi-Objective Reinforcement Learning (MORL) to achieve Preference-Conditioned Humanoid Control (PCHC). Unlike conventional methods that require training a series of policies to approximate the Pareto front, our framework enables a single, preference-conditioned policy to exhibit a wide spectrum of diverse behaviors. To effectively integrate these requirements, we introduce a Beta distribution-based alignment mechanism based on preference vectors modulating a Mixture-of-Experts (MoE) module. We validated our approach on two representative humanoid tasks. Extensive simulations and real-world experiments demonstrate that the proposed framework allows the robot to adaptively shift its objective priorities in real-time based on the input preference condition.

Huanyu Li, Kun Lei, Sheng Zang et al.

Post-training algorithms based on deep reinforcement learning can push the limits of robotic models for specific objectives, such as generalizability, accuracy, and robustness. However, Intervention-requiring Failures (IR Failures) (e.g., a robot spilling water or breaking fragile glass) during real-world exploration happen inevitably, hindering the practical deployment of such a paradigm. To tackle this, we introduce Failure-Aware Offline-to-Online Reinforcement Learning (FARL), a new paradigm minimizing failures during real-world reinforcement learning. We create FailureBench, a benchmark that incorporates common failure scenarios requiring human intervention, and propose an algorithm that integrates a world-model-based safety critic and a recovery policy trained offline to prevent failures during online exploration. Extensive simulation and real-world experiments demonstrate the effectiveness of FARL in significantly reducing IR Failures while improving performance and generalization during online reinforcement learning post-training. FARL reduces IR Failures by 73.1% while elevating performance by 11.3% on average during real-world RL post-training. Videos and code are available at https://failure-aware-rl.github.io.

Huanyu Li, Zongyuan Li, Wei Huang et al.

Large language models (LLMs) such as ChatGPT o1, ChatGPT o3, and DeepSeek R1 have shown great potential in solving difficult problems. However, current LLM evaluation benchmarks are limited to one-step interactions. Some of the existing sequence decision-making environments, such as TextStarCraftII and LLM-PySC2, are too complicated and require hours of interaction to complete a game. In this paper, we introduce LLM-Cave, a benchmark and light environment for LLM reasoning and decision-making systems. This environment is a classic instance in the era of Symbolism. Artificial intelligence enables the agent to explore the environment and avoid potential losses by reasoning about nearby dangers using partial observable state information. In the experiment, we evaluated the sequential reasoning ability, decision-making performance and computational efficiency of mainstream large language models (LLMs) such as GPT-4o-mini, o1-mini, and DeepSeek-R1. Experiments show that while Deepseek-R1 achieved the highest success rate on complex reasoning tasks, smaller models like 4o-mini significantly narrowed the performance gap on challenges by employing Chain of Speculation and Planner-Critic strategies, at the expense of reduced computational efficiency. This indicates that structured, multi-step reasoning combined with an LLM-based feedback mechanism can substantially enhance an LLM's decision-making capabilities, providing a promising direction for improving reasoning in weaker models and suggesting a new reasoning-centered benchmark for LLM assessment. Our code is open-sourced in https://github.com/puleya1277/CaveEnv.

Weiji Xie, Jinrui Han, Jiakun Zheng et al.

Humanoid robots are promising to acquire various skills by imitating human behaviors. However, existing algorithms are only capable of tracking smooth, low-speed human motions, even with delicate reward and curriculum design. This paper presents a physics-based humanoid control framework, aiming to master highly-dynamic human behaviors such as Kungfu and dancing through multi-steps motion processing and adaptive motion tracking. For motion processing, we design a pipeline to extract, filter out, correct, and retarget motions, while ensuring compliance with physical constraints to the maximum extent. For motion imitation, we formulate a bi-level optimization problem to dynamically adjust the tracking accuracy tolerance based on the current tracking error, creating an adaptive curriculum mechanism. We further construct an asymmetric actor-critic framework for policy training. In experiments, we train whole-body control policies to imitate a set of highly-dynamic motions. Our method achieves significantly lower tracking errors than existing approaches and is successfully deployed on the Unitree G1 robot, demonstrating stable and expressive behaviors. The project page is https://kungfu-bot.github.io.

Guangqi Jiang, Yifei Sun, Tao Huang et al.

The pre-training of visual representations has enhanced the efficiency of robot learning. Due to the lack of large-scale in-domain robotic datasets, prior works utilize in-the-wild human videos to pre-train robotic visual representation. Despite their promising results, representations from human videos are inevitably subject to distribution shifts and lack the dynamics information crucial for task completion. We first evaluate various pre-trained representations in terms of their correlation to the downstream robotic manipulation tasks (i.e., manipulation centricity). Interestingly, we find that the "manipulation centricity" is a strong indicator of success rates when applied to downstream tasks. Drawing from these findings, we propose Manipulation Centric Representation (MCR), a foundation representation learning framework capturing both visual features and the dynamics information such as actions and proprioceptions of manipulation tasks to improve manipulation centricity. Specifically, we pre-train a visual encoder on the DROID robotic dataset and leverage motion-relevant data such as robot proprioceptive states and actions. We introduce a novel contrastive loss that aligns visual observations with the robot's proprioceptive state-action dynamics, combined with a behavior cloning (BC)-like actor loss to predict actions during pre-training, along with a time contrastive loss. Empirical results across 4 simulation domains with 20 tasks verify that MCR outperforms the strongest baseline method by 14.8%. Moreover, MCR boosts the performance of data-efficient learning with a UR5e arm on 3 real-world tasks by 76.9%. Project website: https://robots-pretrain-robots.github.io/.

Zongyuan Li, Yanan Ni, Runnan Qi et al.

The tremendous potential has been demonstrated by large language models (LLMs) in intelligent decision-making problems, with unprecedented capabilities shown across diverse applications ranging from gaming AI systems to complex strategic planning frameworks. However, the StarCraft II platform, which has been widely adopted for validating decision-making algorithms in the past decade, has not yet provided substantial support for this emerging domain. To address issues that LLMs cannot interface with the hundreds of actions of the pysc2 backend and the lack of native support for multi-agent (MA) collaboration, we propose the LLM-PySC2 environment. This is the first environment that offers LLMs the complete pysc2 action space with sufficient multi-modal information and game Wiki knowledge. With an asynchronous query architecture, the environment efficiently interacts with LLMs that maintain a constant latency regardless of the scale of the agents' population. In the experiments, we evaluated LLMs' decision-making performance in both the macro-decision and micro-operation scenarios, with traditional StarCraft II Multi-Agent Challenge (SMAC) tasks and a series of new proposed. Results indicate that LLMs possess the potential to achieve victories in complex scenarios but cannot constantly generate correct decisions, especially in the recovered pysc2 action space and MA settings. Without task-relevant instructions, the pre-trained models suffer from issues such as hallucinations and inefficient collaboration. Our findings suggest that StarCraft II still challenges in the era of large models, revealing that there is a lot to do to develop an advanced LLM decision-making system, and the proposed LLM-PySC2 environment will support future development of LLM-based decision-making solutions.

Kun Lei, Huanyu Li, Dongjie Yu et al.

Real-world robotic manipulation in homes and factories demands reliability, efficiency, and robustness that approach or surpass the performance of skilled human operators. We present RL-100, a real-world reinforcement learning framework built on diffusion-based visuomotor policies. RL-100 unifies imitation and reinforcement learning under a single PPO-style objective applied within the denoising process, yielding conservative and stable policy improvements across both offline and online stages. To meet deployment latency constraints, we employ a lightweight consistency distillation procedure that compresses multi-step diffusion into a one-step controller for high-frequency control. The framework is task-, embodiment-, and representation-agnostic, and supports both single-action outputs and action-chunking control. We evaluate RL-100 on seven diverse real-robot manipulation tasks, ranging from dynamic pushing and agile bowling to pouring, cloth folding, unscrewing, and multi-stage juicing. RL-100 attains 100% success across evaluated trials, achieving 900 out of 900 successful episodes, including up to 250 out of 250 consecutive trials on one task, and matches or surpasses expert teleoperators in time-to-completion. Without retraining, a single policy attains approximately 90% zero-shot success under environmental and dynamics shifts, adapts in a few-shot regime to significant task variations (86.7%), and remains robust to aggressive human perturbations (about 95%). In a public shopping-mall deployment, the juicing robot served random customers continuously for roughly seven hours without failure. Together, these results suggest a practical path toward deployment-ready robot learning: start from human priors, align training objectives with human-grounded metrics, and reliably extend performance beyond human demonstrations.

Jianda Yue, Tingting Li, Jian Ouyang et al.

Taste peptides have emerged as promising natural flavoring agents attributed to their unique organoleptic properties, high safety profile, and potential health benefits. However, the de novo identification of taste peptides derived from animal, plant, or microbial sources remains a time-consuming and resource-intensive process, significantly impeding their widespread application in the food industry. Here, we present TastePepAI, a comprehensive artificial intelligence framework for customized taste peptide design and safety assessment. As the key element of this framework, a loss-supervised adaptive variational autoencoder (LA-VAE) is implemented to efficiently optimizes the latent representation of sequences during training and facilitates the generation of target peptides with desired taste profiles. Notably, our model incorporates a novel taste-avoidance mechanism, allowing for selective flavor exclusion. Subsequently, our in-house developed toxicity prediction algorithm (SpepToxPred) is integrated in the framework to undergo rigorous safety evaluation of generated peptides. Using this integrated platform, we successfully identified 73 peptides exhibiting sweet, salty, and umami, significantly expanding the current repertoire of taste peptides. This work demonstrates the potential of TastePepAI in accelerating taste peptide discovery for food applications and provides a versatile framework adaptable to broader peptide engineering challenges.

Huanyu Li, Rickard Armiento, Patrick Lambrix

In the materials design domain, much of the data from materials calculations are stored in different heterogeneous databases. Materials databases usually have different data models. Therefore, the users have to face the challenges to find the data from adequate sources and integrate data from multiple sources. Ontologies and ontology-based techniques can address such problems as the formal representation of domain knowledge can make data more available and interoperable among different systems. In this paper, we introduce the Materials Design Ontology (MDO), which defines concepts and relations to cover knowledge in the field of materials design. MDO is designed using domain knowledge in materials science (especially in solid-state physics), and is guided by the data from several databases in the materials design field. We show the application of the MDO to materials data retrieved from well-known materials databases.