RLBayes: a Bayesian Network Structure Learning Algorithm via Reinforcement Learning-Based Search Strategy

The score-based structure learning of Bayesian network (BN) is an effective way to learn BN models, which are regarded as some of the most compelling probabilistic graphical models in the field of representation and reasoning under uncertainty. However, the search space of structure learning grows super-exponentially as the number of variables increases, which makes BN structure learning an NP-hard problem, as well as a combination optimization problem (COP). Despite the successes of many heuristic methods on it, the results of the structure learning of BN are usually unsatisfactory. Inspired by Q-learning, in this paper, a Bayesian network structure learning algorithm via reinforcement learning-based (RL-based) search strategy is proposed, namely RLBayes. The method borrows the idea of RL and tends to record and guide the learning process by a dynamically maintained Q-table. By creating and maintaining the dynamic Q-table, RLBayes achieve storing the unlimited search space within limited space, thereby achieving the structure learning of BN via Q-learning. Not only is it theoretically proved that RLBayes can converge to the global optimal BN structure, but also it is experimentally proved that RLBayes has a better effect than almost all other heuristic search algorithms.

A Comprehensive Analysis of Mamba for 3D Volumetric Medical Image Segmentation

Mamba, with its selective State Space Models (SSMs), offers a more computationally efficient solution than Transformers for long-range dependency modeling. However, there is still a debate about its effectiveness in high-resolution 3D medical image segmentation. In this study, we present a comprehensive investigation into Mamba's capabilities in 3D medical image segmentation by tackling three pivotal questions: Can Mamba replace Transformers? Can it elevate multi-scale representation learning? Is complex scanning necessary to unlock its full potential? We evaluate Mamba's performance across three large public benchmarks-AMOS, TotalSegmentator, and BraTS. Our findings reveal that UlikeMamba, a U-shape Mamba-based network, consistently surpasses UlikeTrans, a U-shape Transformer-based network, particularly when enhanced with custom-designed 3D depthwise convolutions, boosting accuracy and computational efficiency. Further, our proposed multi-scale Mamba block demonstrates superior performance in capturing both fine-grained details and global context, especially in complex segmentation tasks, surpassing Transformer-based counterparts. We also critically assess complex scanning strategies, finding that simpler methods often suffice, while our Tri-scan approach delivers notable advantages in the most challenging scenarios. By integrating these advancements, we introduce a new network for 3D medical image segmentation, positioning Mamba as a transformative force that outperforms leading models such as nnUNet, CoTr, and U-Mamba, offering competitive accuracy with superior computational efficiency. This study provides key insights into Mamba's unique advantages, paving the way for more efficient and accurate approaches to 3D medical imaging.

Towards Long-Range ENSO Prediction with an Explainable Deep Learning Model

El Ni\~no-Southern Oscillation (ENSO) is a prominent mode of interannual climate variability with far-reaching global impacts. Its evolution is governed by intricate air-sea interactions, posing significant challenges for long-term prediction. In this study, we introduce CTEFNet, a multivariate deep learning model that synergizes convolutional neural networks and transformers to enhance ENSO forecasting. By integrating multiple oceanic and atmospheric predictors, CTEFNet extends the effective forecast lead time to 20 months while mitigating the impact of the spring predictability barrier, outperforming both dynamical models and state-of-the-art deep learning approaches. Furthermore, CTEFNet offers physically meaningful and statistically significant insights through gradient-based sensitivity analysis, revealing the key precursor signals that govern ENSO dynamics, which align with well-established theories and reveal new insights about inter-basin interactions among the Pacific, Atlantic, and Indian Oceans. The CTEFNet's superior predictive skill and interpretable sensitivity assessments underscore its potential for advancing climate prediction. Our findings highlight the importance of multivariate coupling in ENSO evolution and demonstrate the promise of deep learning in capturing complex climate dynamics with enhanced interpretability.

A2DO: Adaptive Anti-Degradation Odometry with Deep Multi-Sensor Fusion for Autonomous Navigation

Accurate localization is essential for the safe and effective navigation of autonomous vehicles, and Simultaneous Localization and Mapping (SLAM) is a cornerstone technology in this context. However, The performance of the SLAM system can deteriorate under challenging conditions such as low light, adverse weather, or obstructions due to sensor degradation. We present A2DO, a novel end-to-end multi-sensor fusion odometry system that enhances robustness in these scenarios through deep neural networks. A2DO integrates LiDAR and visual data, employing a multi-layer, multi-scale feature encoding module augmented by an attention mechanism to mitigate sensor degradation dynamically. The system is pre-trained extensively on simulated datasets covering a broad range of degradation scenarios and fine-tuned on a curated set of real-world data, ensuring robust adaptation to complex scenarios. Our experiments demonstrate that A2DO maintains superior localization accuracy and robustness across various degradation conditions, showcasing its potential for practical implementation in autonomous vehicle systems.

Efficient Response Generation Method Selection for Fine-Tuning Large Language Models

The training data for fine-tuning large language models (LLMs) is typically structured as input-output pairs. However, for many tasks, there can be multiple equally valid output variations for the same input. Recent studies have observed that the choice of output variation used in training can affect the model's performance. This raises an important question: how can we generate the most effective output from the many possible response generation strategy options? Rather than relying on the traditional but resource-intensive train-and-evaluate approach, this paper proposes a scalable, approximate method for estimating the quality of a small subset of generated training data derived from the same input. We then evaluate how well this small subset of generated output fits the target model we are trying to train. We present a large-scale benchmark covering diverse reasoning-based datasets to support our study. The central idea is that a good output should closely resemble the output generated by the target LLM. We formalize this 'closeness' as the expected alignment score between a candidate output and the output sampled from the target LLM. We connect this measurement to the perplexity metric used in previous literature and demonstrate that leveraging an alignment-based metric can provide better predictions of model performance. Using this strategy, we can evaluate a small subset of the generated output from each response generation strategy option, then select the most effective strategy. We show that an LLM trained on data generated by the selected strategy could lead to a significant performance gain in many cases.

Beyond Prompt Content: Enhancing LLM Performance via Content-Format Integrated Prompt Optimization

Large Language Models (LLMs) have shown significant capability across various tasks, with their real-world effectiveness often driven by prompt design. While recent research has focused on optimizing prompt content, the role of prompt formatting, a critical but often overlooked dimension, has received limited systematic investigation. In this paper, we introduce Content-Format Integrated Prompt Optimization (CFPO), an innovative methodology that jointly optimizes both prompt content and formatting through an iterative refinement process. CFPO leverages natural language mutations to explore content variations and employs a dynamic format exploration strategy that systematically evaluates diverse format options. Our extensive evaluations across multiple tasks and open-source LLMs demonstrate that CFPO demonstrates measurable performance improvements compared to content-only optimization methods. This highlights the importance of integrated content-format optimization and offers a practical, model-agnostic approach to enhancing LLM performance. Code is available at https://github.com/HenryLau7/CFPO.

MADP: Multi-Agent Deductive Planning for Enhanced Cognitive-Behavioral Mental Health Question Answer

The Mental Health Question Answer (MHQA) task requires the seeker and supporter to complete the support process in one-turn dialogue. Given the richness of help-seeker posts, supporters must thoroughly understand the content and provide logical, comprehensive, and well-structured responses. Previous works in MHQA mostly focus on single-agent approaches based on the cognitive element of Cognitive Behavioral Therapy (CBT), but they overlook the interactions among various CBT elements, such as emotion and cognition. This limitation hinders the models' ability to thoroughly understand the distress of help-seekers. To address this, we propose a framework named Multi-Agent Deductive Planning (MADP), which is based on the interactions between the various psychological elements of CBT. This method guides Large Language Models (LLMs) to achieve a deeper understanding of the seeker's context and provide more personalized assistance based on individual circumstances. Furthermore, we construct a new dataset based on the MADP framework and use it to fine-tune LLMs, resulting in a specialized model named MADP-LLM. We conduct extensive experiments, including comparisons with multiple LLMs, human evaluations, and automatic evaluations, to validate the effectiveness of the MADP framework and MADP-LLM.

Sigma: Differential Rescaling of Query, Key and Value for Efficient Language Models

We introduce Sigma, an efficient large language model specialized for the system domain, empowered by a novel architecture including DiffQKV attention, and pre-trained on our meticulously collected system domain data. DiffQKV attention significantly enhances the inference efficiency of Sigma by optimizing the Query (Q), Key (K), and Value (V) components in the attention mechanism differentially, based on their varying impacts on the model performance and efficiency indicators. Specifically, we (1) conduct extensive experiments that demonstrate the model's varying sensitivity to the compression of K and V components, leading to the development of differentially compressed KV, and (2) propose augmented Q to expand the Q head dimension, which enhances the model's representation capacity with minimal impacts on the inference speed. Rigorous theoretical and empirical analyses reveal that DiffQKV attention significantly enhances efficiency, achieving up to a 33.36% improvement in inference speed over the conventional grouped-query attention (GQA) in long-context scenarios. We pre-train Sigma on 6T tokens from various sources, including 19.5B system domain data that we carefully collect and 1T tokens of synthesized and rewritten data. In general domains, Sigma achieves comparable performance to other state-of-arts models. In the system domain, we introduce the first comprehensive benchmark AIMicius, where Sigma demonstrates remarkable performance across all tasks, significantly outperforming GPT-4 with an absolute improvement up to 52.5%.

IPDN: Image-enhanced Prompt Decoding Network for 3D Referring Expression Segmentation

3D Referring Expression Segmentation (3D-RES) aims to segment point cloud scenes based on a given expression. However, existing 3D-RES approaches face two major challenges: feature ambiguity and intent ambiguity. Feature ambiguity arises from information loss or distortion during point cloud acquisition due to limitations such as lighting and viewpoint. Intent ambiguity refers to the model's equal treatment of all queries during the decoding process, lacking top-down task-specific guidance. In this paper, we introduce an Image enhanced Prompt Decoding Network (IPDN), which leverages multi-view images and task-driven information to enhance the model's reasoning capabilities. To address feature ambiguity, we propose the Multi-view Semantic Embedding (MSE) module, which injects multi-view 2D image information into the 3D scene and compensates for potential spatial information loss. To tackle intent ambiguity, we designed a Prompt-Aware Decoder (PAD) that guides the decoding process by deriving task-driven signals from the interaction between the expression and visual features. Comprehensive experiments demonstrate that IPDN outperforms the state-ofthe-art by 1.9 and 4.2 points in mIoU metrics on the 3D-RES and 3D-GRES tasks, respectively.

EpiCoder: Encompassing Diversity and Complexity in Code Generation

Effective instruction tuning is indispensable for optimizing code LLMs, aligning model behavior with user expectations and enhancing model performance in real-world applications. However, most existing methods focus on code snippets, which are limited to specific functionalities and rigid structures, restricting the complexity and diversity of the synthesized data. To address these limitations, we introduce a novel feature tree-based synthesis framework inspired by Abstract Syntax Trees (AST). Unlike AST, which captures syntactic structure of code, our framework models semantic relationships between code elements, enabling the generation of more nuanced and diverse data. The feature tree is constructed from raw data and refined iteratively to increase the quantity and diversity of the extracted features. This process enables the identification of more complex patterns and relationships within the code. By sampling subtrees with controlled depth and breadth, our framework allows precise adjustments to the complexity of the generated code, supporting a wide range of tasks from simple function-level operations to intricate multi-file scenarios. We fine-tuned widely-used base models to create the EpiCoder series, achieving state-of-the-art performance at both the function and file levels across multiple benchmarks. Notably, empirical evidence indicates that our approach shows significant potential in synthesizing highly complex repository-level code data. Further analysis elucidates the merits of this approach by rigorously assessing data complexity and diversity through software engineering principles and LLM-as-a-judge method.