Wenlong Wang

Wenlong Wang, Ivana Dusparic, Yucheng Shi et al.

Model-based reinforcement learning (RL) offers a solution to the data inefficiency that plagues most model-free RL algorithms. However, learning a robust world model often requires complex and deep architectures, which are computationally expensive and challenging to train. Within the world model, sequence models play a critical role in accurate predictions, and various architectures have been explored, each with its own challenges. Currently, recurrent neural network (RNN)-based world models struggle with vanishing gradients and capturing long-term dependencies. Transformers, on the other hand, suffer from the quadratic memory and computational complexity of self-attention mechanisms, scaling as $O(n^2)$, where $n$ is the sequence length. To address these challenges, we propose a state space model (SSM)-based world model, Drama, specifically leveraging Mamba, that achieves $O(n)$ memory and computational complexity while effectively capturing long-term dependencies and enabling efficient training with longer sequences. We also introduce a novel sampling method to mitigate the suboptimality caused by an incorrect world model in the early training stages. Combining these techniques, Drama achieves a normalised score on the Atari100k benchmark that is competitive with other state-of-the-art (SOTA) model-based RL algorithms, using only a 7 million-parameter world model. Drama is accessible and trainable on off-the-shelf hardware, such as a standard laptop. Our code is available at https://github.com/realwenlongwang/Drama.git.

Davide Bianchi, Florian Bossmann, Wenlong Wang et al.

We introduce a data-adaptive inversion method that integrates classical or deep learning-based approaches with iterative graph Laplacian regularization, specifically targeting acoustic impedance inversion - a critical task in seismic exploration. Our method initiates from an impedance estimate derived using either traditional inversion techniques or neural network-based methods. This initial estimate guides the construction of a graph Laplacian operator, effectively capturing structural characteristics of the impedance profile. Utilizing a Tikhonov-inspired variational framework with this graph-informed prior, our approach iteratively updates and refines the impedance estimate while continuously recalibrating the graph Laplacian. This iterative refinement shows rapid convergence, increased accuracy, and enhanced robustness to noise compared to initial reconstructions alone. Extensive validation performed on synthetic and real seismic datasets across varying noise levels confirms the effectiveness of our method. Performance evaluations include four initial inversion methods: two classical techniques and two neural networks - previously established in the literature.

Wenlong Wang, Fergal Reid

Recent work on recursive reasoning models like TRM demonstrates that tiny networks (7M parameters) can achieve strong performance on abstract reasoning tasks through latent recursion -- iterative refinement in hidden representation space without emitting intermediate tokens. This raises a natural question about operator choice: Mamba-2's state space recurrence is itself a form of iterative refinement, making it a natural candidate for recursive reasoning -- but does introducing Mamba-2 into the recursive scaffold preserve reasoning capability? We investigate this by replacing the Transformer blocks in TRM with Mamba-2 hybrid operators while maintaining parameter parity (6.83M vs 6.86M parameters). On ARC-AGI-1, we find that the hybrid improves pass@2 (the official metric) by +2.0\% (45.88\% vs 43.88\%) and consistently outperforms at higher K values (+4.75\% at pass@100), whilst maintaining pass@1 parity. This suggests improved candidate coverage -- the model generates correct solutions more reliably -- with similar top-1 selection. Our results validate that Mamba-2 hybrid operators preserve reasoning capability within the recursive scaffold, establishing SSM-based operators as viable candidates in the recursive operator design space and taking a first step towards understanding the best mixing strategies for recursive reasoning.

Weihua Cheng, Ersheng Ni, Wenlong Wang et al.

The rapid progress of Large Language Models (LLMs) and their multimodal extensions (MLLMs) has enabled agentic systems capable of perceiving and acting across diverse environments. A challenging yet impactful frontier is the development of GUI agents, which must navigate complex desktop and web interfaces while maintaining robustness and generalization. Existing paradigms typically model tasks as long-chain executions, concatenating historical trajectories into the context. While approaches such as Mirage and GTA1 refine planning or introduce multi-branch action selection, they remain constrained by two persistent issues: Dependence on historical trajectories, which amplifies error propagation. And Local exploration bias, where "decision-first, observation-later" mechanisms overlook critical interface cues. We introduce the Memory-Driven GUI Agent (MGA), which reframes GUI interaction around the principle of observe first, then decide. MGA models each step as an independent, context-rich environment state represented by a triad: current screenshot, task-agnostic spatial information, and a dynamically updated structured memory. Experiments on OSworld benchmarks, real desktop applications (Chrome, VSCode, VLC), and cross-task transfer demonstrate that MGA achieves substantial gains in robustness, generalization, and efficiency compared to state-of-the-art baselines. The code is publicly available at: {https://anonymous.4open.science/r/MGA-3571}.

Yulin Sun, Xiaopeng Si, Runnan He et al.

Timely identification of harmful brain activities via electroencephalography (EEG) is critical for brain disease diagnosis and treatment, which remains limited application due to inter-rater variability, resource constraints, and poor generalizability of existing artificial intelligence (AI) models. In this study, a convolutional neural network model, VIPEEGNet, was developed and validated using EEGs recorded from Massachusetts General Hospital/Harvard Medical School. The VIPEEGNet was developed and validated using two independent datasets, collected between 2006 and 2020. The development cohort included EEG recordings from 1950 patients, with 106,800 EEG segments annotated by at least one experts (ranging from 1 to 28). The online testing cohort consisted of EEG segments from a subset of an additional 1,532 patients, each annotated by at least 10 experts. For the development cohort (n=1950), the VIPEEGNet achieved high accuracy, with an AUROC for binary classification of seizure, LPD, GPD, LRDA, GRDA, and "other" categories at 0.972 (95% CI, 0.957-0.988), 0.962 (95% CI, 0.954-0.970), 0.972 (95% CI, 0.960-0.984), 0.938 (95% CI, 0.917-0.959), 0.949 (95% CI, 0.941-0.957), and 0.930 (95% CI, 0.926-0.935). For multi classification, the sensitivity of VIPEEGNET for the six categories ranges from 36.8% to 88.2% and the precision ranges from 55.6% to 80.4%, and performance similar to human experts. Notably, the external validation showed Kullback-Leibler Divergence (KLD)of 0.223 and 0.273, ranking top 2 among the existing 2,767 competing algorithms, while we only used 2.8% of the parameters of the first-ranked algorithm.

Yucheng Shi, Wenlong Wang, Xiaowen Tao et al.

Dynamic scheduling of access to shared resources by autonomous systems is a challenging problem, characterized as being NP-hard. The complexity of this task leads to a combinatorial explosion of possibilities in highly dynamic systems where arriving requests must be continuously scheduled subject to strong safety and time constraints. An example of such a system is an unsignalized intersection, where automated vehicles' access to potential conflict zones must be dynamically scheduled. In this paper, we apply Neural Monte Carlo Tree Search (NMCTS) to the challenging task of scheduling platoons of vehicles crossing unsignalized intersections. Crucially, we introduce a transformation model that maps successive sequences of potentially conflicting road-space reservation requests from platoons of vehicles into a series of board-game-like problems and use NMCTS to search for solutions representing optimal road-space allocation schedules in the context of past allocations. To optimize search, we incorporate a prioritized re-sampling method with parallel NMCTS (PNMCTS) to improve the quality of training data. To optimize training, a curriculum learning strategy is used to train the agent to schedule progressively more complex boards culminating in overlapping boards that represent busy intersections. In a busy single four-way unsignalized intersection simulation, PNMCTS solved 95\% of unseen scenarios, reducing crossing time by 43\% in light and 52\% in heavy traffic versus first-in, first-out control. In a 3x3 multi-intersection network, the proposed method maintained free-flow in light traffic when all intersections are under control of PNMCTS and outperformed state-of-the-art RL-based traffic-light controllers in average travel time by 74.5\% and total throughput by 16\% in heavy traffic.

Wenlong Wang, David Hung-Chang Du

We present LearnedKV, a novel tiered key-value store that seamlessly integrates a Log-Structured Merge (LSM) tree with a Learned Index to achieve superior read and write performance on storage systems. While existing approaches use learned indexes primarily as auxiliary components within LSM trees, LearnedKV employs a two-tier design where the LSM tree handles recent write operations while a separate Learned Index accelerates read performance. Our design includes a non-blocking conversion mechanism that efficiently transforms LSM data into a Learned Index during garbage collection, maintaining high performance without interrupting operations. LearnedKV dramatically reduces LSM size through this tiered approach, leading to significant performance gains in both reads and writes. Extensive evaluations across diverse workloads show that LearnedKV outperforms state-of-the-art LSM-based solutions by up to 4.32x for read operations and 1.43x for writes. The system demonstrates robust performance across different data distributions, access patterns, and storage media including both SSDs and HDDs.

Muyang Ge, Wenlong Wang, Wangxiangming Zheng

Seismic impedance inversion can be performed with a semi-supervised learning algorithm, which only needs a few logs as labels and is less likely to get overfitted. However, classical semi-supervised learning algorithm usually leads to artifacts on the predicted impedance image. In this artical, we improve the semi-supervised learning from two aspects. First, by replacing 1-d convolutional neural network (CNN) layers in deep learning structure with 2-d CNN layers and 2-d maxpooling layers, the prediction accuracy is improved. Second, prediction uncertainty can also be estimated by embedding the network into a Bayesian inference framework. Local reparameterization trick is used during forward propagation of the network to reduce sampling cost. Tests with Marmousi2 model and SEAM model validate the feasibility of the proposed strategy.

Siwei Yu, Jianwei Ma, Wenlong Wang

Compared with traditional seismic noise attenuation algorithms that depend on signal models and their corresponding prior assumptions, removing noise with a deep neural network is trained based on a large training set, where the inputs are the raw datasets and the corresponding outputs are the desired clean data. After the completion of training, the deep learning method achieves adaptive denoising with no requirements of (i) accurate modelings of the signal and noise, or (ii) optimal parameters tuning. We call this intelligent denoising. We use a convolutional neural network as the basic tool for deep learning. In random and linear noise attenuation, the training set is generated with artificially added noise. In the multiple attenuation step, the training set is generated with acoustic wave equation. Stochastic gradient descent is used to solve the optimal parameters for the convolutional neural network. The runtime of deep learning on a graphics processing unit for denoising has the same order as the $f-x$ deconvolution method. Synthetic and field results show the potential applications of deep learning in automatic attenuation of random noise (with unknown variance), linear noise, and multiples.