Search

Recent learning-to-imitation methods have shown promising results in planning via imitating within the observation-action space. However, their ability in open environments remains constrained, particularly in long-horizon tasks. In contrast, traditional symbolic planning excels in long-horizon tasks through logical reasoning over human-defined symbolic spaces but struggles to handle observations beyond symbolic states, such as high-dimensional visual inputs encountered in real-world scenarios. In this work, we draw inspiration from abductive learning and introduce a novel framework \textbf{AB}ductive \textbf{I}mitation \textbf{L}earning (ABIL) that integrates the benefits of data-driven learning and symbolic-based reasoning, enabling long-horizon planning. Specifically, we employ abductive reasoning to understand the demonstrations in symbolic space and design the principles of sequential consistency to resolve the conflicts between perception and reasoning. ABIL generates predicate candidates to facilitate the perception from raw observations to symbolic space without laborious predicate annotations, providing a groundwork for symbolic planning. With the symbolic understanding, we further develop a policy ensemble whose base policies are built with different logical objectives and managed through symbolic reasoning. Experiments show that our proposal successfully understands the observations with the task-relevant symbolics to assist the imitation learning. Importantly, ABIL demonstrates significantly improved data efficiency and generalization across various long-horizon tasks, highlighting it as a promising solution for long-horizon planning. Project website: \url{https://www.lamda.nju.edu.cn/shaojj/KDD25_ABIL/}., Comment: Accepted by KDD2025. The KDD version is titled ''Abductive Learning for Neuro-Symbolic Grounded Imitation''

This technical report presents our initial attempt to build a spoken large language model (LLM) for Taiwanese Mandarin, specifically tailored to enable real-time, speech-to-speech interaction in multi-turn conversations. Our end-to-end model incorporates a decoder-only transformer architecture and aims to achieve seamless interaction while preserving the conversational flow, including full-duplex capabilities allowing simultaneous speaking and listening. The paper also details the training process, including data preparation with synthesized dialogues and adjustments for real-time interaction. We also developed a platform to evaluate conversational fluency and response coherence in multi-turn dialogues. We hope the release of the report can contribute to the future development of spoken LLMs in Taiwanese Mandarin., Comment: Work in progress

Multimodal foundation models, such as Gemini and ChatGPT, have revolutionized human-machine interactions by seamlessly integrating various forms of data. Developing a universal spoken language model that comprehends a wide range of natural language instructions is critical for bridging communication gaps and facilitating more intuitive interactions. However, the absence of a comprehensive evaluation benchmark poses a significant challenge. We present Dynamic-SUPERB Phase-2, an open and evolving benchmark for the comprehensive evaluation of instruction-based universal speech models. Building upon the first generation, this second version incorporates 125 new tasks contributed collaboratively by the global research community, expanding the benchmark to a total of 180 tasks, making it the largest benchmark for speech and audio evaluation. While the first generation of Dynamic-SUPERB was limited to classification tasks, Dynamic-SUPERB Phase-2 broadens its evaluation capabilities by introducing a wide array of novel and diverse tasks, including regression and sequence generation, across speech, music, and environmental audio. Evaluation results indicate that none of the models performed well universally. SALMONN-13B excelled in English ASR, while WavLLM demonstrated high accuracy in emotion recognition, but current models still require further innovations to handle a broader range of tasks. We will soon open-source all task data and the evaluation pipeline.

Large Language Models (LLMs) have shown significant potential in designing reward functions for Reinforcement Learning (RL) tasks. However, obtaining high-quality reward code often involves human intervention, numerous LLM queries, or repetitive RL training. To address these issues, we propose CARD, a LLM-driven Reward Design framework that iteratively generates and improves reward function code. Specifically, CARD includes a Coder that generates and verifies the code, while a Evaluator provides dynamic feedback to guide the Coder in improving the code, eliminating the need for human feedback. In addition to process feedback and trajectory feedback, we introduce Trajectory Preference Evaluation (TPE), which evaluates the current reward function based on trajectory preferences. If the code fails the TPE, the Evaluator provides preference feedback, avoiding RL training at every iteration and making the reward function better aligned with the task objective. Empirical results on Meta-World and ManiSkill2 demonstrate that our method achieves an effective balance between task performance and token efficiency, outperforming or matching the baselines across all tasks. On 10 out of 12 tasks, CARD shows better or comparable performance to policies trained with expert-designed rewards, and our method even surpasses the oracle on 3 tasks.

Explainable Recommendation task is designed to receive a pair of user and item and output explanations to justify why an item is recommended to a user. Many models treat review-generation as a proxy of explainable recommendation. Although they are able to generate fluent and grammatical sentences, they suffer from generality and hallucination issues. We propose a personalized, aspect-controlled model called Multi-Aspect Prompt LEarner (MAPLE), in which it integrates aspect category as another input dimension to facilitate the memorization of fine-grained aspect terms. Experiments on two real-world review datasets in restaurant domain show that MAPLE outperforms the baseline review-generation models in terms of text and feature diversity while maintaining excellent coherence and factual relevance. We further treat MAPLE as a retriever component in the retriever-reader framework and employ a Large-Language Model (LLM) as the reader, showing that MAPLE's explanation along with the LLM's comprehension ability leads to enriched and personalized explanation as a result. We will release the code and data in this http upon acceptance., Comment: 8 main pages, 10 pages for appendix. Under review

Large language models (LLMs), including both proprietary and open-source models, have showcased remarkable capabilities in addressing a wide range of downstream tasks. Nonetheless, when it comes to practical Chinese legal tasks, these models fail to meet the actual requirements. Proprietary models do not ensure data privacy for sensitive legal cases, while open-source models demonstrate unsatisfactory performance due to their lack of legal knowledge. To address this problem, we introduce LawGPT, the first open-source model specifically designed for Chinese legal applications. LawGPT comprises two key components: legal-oriented pre-training and legal supervised fine-tuning. Specifically, we employ large-scale Chinese legal documents for legal-oriented pre-training to incorporate legal domain knowledge. To further improve the model's performance on downstream legal tasks, we create a knowledge-driven instruction dataset for legal supervised fine-tuning. Our experimental results demonstrate that LawGPT outperforms the open-source LLaMA 7B model. Our code and resources are publicly available at https://github.com/pengxiao-song/LaWGPT and have received 5.7K stars on GitHub., Comment: Technical Report

The foundation model paradigm leverages a shared foundation model to achieve state-of-the-art (SOTA) performance for various tasks, requiring minimal downstream-specific modeling and data annotation. This approach has proven crucial in the field of Natural Language Processing (NLP). However, the speech processing community lacks a similar setup to explore the paradigm systematically. In this work, we establish the Speech processing Universal PERformance Benchmark (SUPERB) to study the effectiveness of the paradigm for speech. We propose a unified multi-tasking framework to address speech processing tasks in SUPERB using a frozen foundation model followed by task-specialized, lightweight prediction heads. Combining our results with community submissions, we verify that the foundation model paradigm is promising for speech, and our multi-tasking framework is simple yet effective, as the best-performing foundation model shows competitive generalizability across most SUPERB tasks. For reproducibility and extensibility, we have developed a long-term maintained platform that enables deterministic benchmarking, allows for result sharing via an online leaderboard, and promotes collaboration through a community-driven benchmark database to support new development cycles. Finally, we conduct a series of analyses to offer an in-depth understanding of SUPERB and speech foundation models, including information flows across tasks inside the models, the correctness of the weighted-sum benchmarking protocol and the statistical significance and robustness of the benchmark., Comment: The extended journal version for SUPERB and SUPERB-SG. Published in IEEE/ACM TASLP. The Arxiv version is preferred

Surfaces (interfaces) dictate many physical and chemical properties of solid materials and adsorbates considerably affect these properties. Nitrogen molecules, which are the most abundant constituent in ambient air, are considered to be inert. Our study combining atomic force microscopy (AFM), X-ray photoemission spectroscopy (XPS), and thermal desorption spectroscopy (TDS) revealed that nitrogen and water molecules can self-assemble into two-dimensional domains, forming ordered stripe structures on graphitic surfaces in both water and ambient air. The stripe structures of this study were composed of approximately 90% and 10% water and nitrogen molecules, respectively, and survived in ultra-high vacuum (UHV) conditions at temperatures up to approximately 350 K. Because pure water molecules completely desorb from graphitic surfaces in a UHV at temperatures lower than 200 K, our results indicate that the incorporation of nitrogen molecules substantially enhanced the stability of the crystalline water hydrogen bonding network. Additional studies on interfacial gas hydrates can provide deeper insight into the mechanisms underlying formation of gas hydrates.

The detection of GW170817/GRB170817A implied the strong association between short gamma-ray bursts (SGRBs) and binary neutron star (BNS) mergers which produce gravitational waves (GWs). More evidence is needed to confirm the association and reveal the physical processes of BNS mergers. The upcoming High Altitude Detection of Astronomical Radiation (HADAR) experiment, excelling in a wide field of view (FOV) and a large effective area above tens of GeV, is a hope for the prompt detection of very-high-energy (VHE; > 10 GeV) SGRBs. The aim of this paper is to simulate and analyse GW/SGRB joint detections by future GW detector networks in synergy with HADAR, including the second generation LIGO, Virgo and KAGRA and the third generation ET and CE. We provide a brief introduction of the HADAR experiment for SGRB simulations and its expected SGRB detections. For GW simulations, we adopt a phenomenological model to describe GWs produced by BNS mergers and introduce the signal-noise ratios (SNRs) as detector responses. Following a theoretical analysis we compute the redshift-dependent efficiency functions of GW detector networks. We then construct the simulation of GW detection by Monte Carlo sampling. We compare the simulated results of LIGO-Virgo O2 and O3 runs with their actual detections as a check. The combination of GW and SGRB models is then discussed for joint detection, including parameter correlations, triggered SNRs and efficiency skymaps. The estimated joint detection rates are 0.09-2.52 per year for LHVK network with HADAR under different possible configurations, and approximately 0.27-7.89 per year for ET+CE network with HADAR.

This study introduces an innovative 3D printed dry electrode tailored for biosensing in postoperative recovery scenarios. Fabricated through a drop coating process, the electrode incorporates a novel 2D material., Comment: the abstract can't uploaded fully

Contrastive Language-Image Pre-training (CLIP) provides a foundation model by integrating natural language into visual concepts, enabling zero-shot recognition on downstream tasks. It is usually expected that satisfactory overall accuracy can be achieved across numerous domains through well-designed textual prompts. However, we found that their performance in the worst categories is significantly inferior to the overall performance. For example, on ImageNet, there are a total of 10 categories with class-wise accuracy as low as 0\%, even though the overall performance has achieved 64.1\%. This phenomenon reveals the potential risks associated with using CLIP models, particularly in risk-sensitive applications where specific categories hold significant importance. To address this issue, we investigate the alignment between the two modalities in the CLIP model and propose the Class-wise Matching Margin (\cmm) to measure the inference confusion. \cmm\ can effectively identify the worst-performing categories and estimate the potential performance of the candidate prompts. We further query large language models to enrich descriptions of worst-performing categories and build a weighted ensemble to highlight the efficient prompts. Experimental results clearly verify the effectiveness of our proposal, where the accuracy on the worst-10 categories on ImageNet is boosted to 5.2\%, without manual prompt engineering, laborious optimization, or access to labeled validation data.

We investigate the mesoscopic transport through a twisted bilayer graphene (TBG) consisting of a clean graphene nanoribbon on the bottom and a disordered graphene disc on the top. We show that, with strong top-layer disorder the transmission through such a device shows a sequence of resonant peaks with respect to the rotation angle $\theta$, where at the resonance angles $\theta_c$ the disc region contains one giant hexagonal moir{\'e} supercell. A further investigation shows that the value of $\theta_c$ shows negligible dependence on the disorder strength, the Fermi energy, and the shape distortion, indicating the resonance is a robust geometric feature of the moir{\'e} supercell. We explain this geometric resonance based on the bound states formed inside the moir{\'e} supercell, with their averaged local density of states dominating at the AA stacking region while minimizing at the AB stacking region. By increasing the interlayer distance, the peak becomes less pronounced which further confirms the role of interlayer coupling. The results presented here suggest a new mechanism to tune the quantum transport signal through the twist angle in disordered moir{\'e} systems.