Chao Xue

Chao Xue, Wei Liu, Shuai Xie et al.

Automated machine learning (AutoML) seeks to build ML models with minimal human effort. While considerable research has been conducted in the area of AutoML in general, aiming to take humans out of the loop when building artificial intelligence (AI) applications, scant literature has focused on how AutoML works well in open-environment scenarios such as the process of training and updating large models, industrial supply chains or the industrial metaverse, where people often face open-loop problems during the search process: they must continuously collect data, update data and models, satisfy the requirements of the development and deployment environment, support massive devices, modify evaluation metrics, etc. Addressing the open-environment issue with pure data-driven approaches requires considerable data, computing resources, and effort from dedicated data engineers, making current AutoML systems and platforms inefficient and computationally intractable. Human-computer interaction is a practical and feasible way to tackle the problem of open-environment AI. In this paper, we introduce OmniForce, a human-centered AutoML (HAML) system that yields both human-assisted ML and ML-assisted human techniques, to put an AutoML system into practice and build adaptive AI in open-environment scenarios. Specifically, we present OmniForce in terms of ML version management; pipeline-driven development and deployment collaborations; a flexible search strategy framework; and widely provisioned and crowdsourced application algorithms, including large models. Furthermore, the (large) models constructed by OmniForce can be automatically turned into remote services in a few minutes; this process is dubbed model as a service (MaaS). Experimental results obtained in multiple search spaces and real-world use cases demonstrate the efficacy and efficiency of OmniForce.

Aichen Cai, Anmeng Zhang, Anyu Li et al.

We introduce JoyAI-LLM Flash, an efficient Mixture-of-Experts (MoE) language model designed to redefine the trade-off between strong performance and token efficiency in the sub-50B parameter regime. JoyAI-LLM Flash is pretrained on a massive corpus of 20 trillion tokens and further optimized through a rigorous post-training pipeline, including supervised fine-tuning (SFT), Direct Preference Optimization (DPO), and large-scale reinforcement learning (RL) across diverse environments. To improve token efficiency, JoyAI-LLM Flash strategically balances \emph{thinking} and \emph{non-thinking} cognitive modes and introduces FiberPO, a novel RL algorithm inspired by fibration theory that decomposes trust-region maintenance into global and local components, providing unified multi-scale stability control for LLM policy optimization. To enhance architectural sparsity, the model comprises 48B total parameters while activating only 2.7B parameters per forward pass, achieving a substantially higher sparsity ratio than contemporary industry leading models of comparable scale. To further improve inference throughput, we adopt a joint training-inference co-design that incorporates dense Multi-Token Prediction (MTP) and Quantization-Aware Training (QAT). We release the checkpoints for both JoyAI-LLM-48B-A3B Base and its post-trained variants on Hugging Face to support the open-source community.

Chao Xue, Di Liang, Sirui Wang et al.

Transformer-based pre-trained models have achieved great improvements in semantic matching. However, existing models still suffer from insufficient ability to capture subtle differences. The modification, addition and deletion of words in sentence pairs may make it difficult for the model to predict their relationship. To alleviate this problem, we propose a novel Dual Path Modeling Framework to enhance the model's ability to perceive subtle differences in sentence pairs by separately modeling affinity and difference semantics. Based on dual-path modeling framework we design the Dual Path Modeling Network (DPM-Net) to recognize semantic relations. And we conduct extensive experiments on 10 well-studied semantic matching and robustness test datasets, and the experimental results show that our proposed method achieves consistent improvements over baselines.

Yifan Duan, Siyuan Zheng, Lihuan Li et al.

Open datasets and benchmarks for entity-level carbon-emission prediction remain fragmented across access, scale, granularity, and evaluation. We introduce GHGbench, an open dataset and benchmark for company- and building-level greenhouse-gas prediction. The company track contains 32,000+ company-year records from 12,000+ firms with Scope 1+2 and Scope 3 disclosures and financial/sectoral signals; the building track harmonises 491,591 building-year records from 13 open sources into a single schema across 26 metropolitan areas (10 U.S., 15 Australian, 1 Singaporean), with climate covariates and multimodal remote-sensing embeddings. GHGbench defines canonical splits with in-distribution and cross-region/city transfer as primary tasks and temporal hold-out plus short-horizon forecasting as supplementary appendix evidence; headline baselines span gradient-boosted trees, a tabular foundation model, MLP, FT-Transformer, and multimodal fusion, with an LLM panel as auxiliary, all evaluated under multi-seed paired-bootstrap tests. Three benchmark-level findings emerge: (i) building emissions are structurally harder than company emissions; (ii) the in-distribution to out-of-distribution gap dwarfs any within-model gap across both the company track and the building track, and a tabular foundation model is, to our knowledge, the first baseline to open a paired-bootstrap-significant gap over tuned trees on a multi-city building-emissions task; (iii) multimodal remote-sensing embeddings help precisely where tabular generalisation breaks. GHGbench also exposes catastrophic city transfer and the sector-factor lookup ceiling as systematic failure modes. Code and reconstruction recipes are available at GHGbench.

Ziqiang Li, Chaoyue Wang, Xue Rui et al.

Few-shot image generation aims to train generative models using a small number of training images. When there are few images available for training (e.g. 10 images), Learning From Scratch (LFS) methods often generate images that closely resemble the training data while Transfer Learning (TL) methods try to improve performance by leveraging prior knowledge from GANs pre-trained on large-scale datasets. However, current TL methods may not allow for sufficient control over the degree of knowledge preservation from the source model, making them unsuitable for setups where the source and target domains are not closely related. To address this, we propose a novel pipeline called Peer is your Pillar (PIP), which combines a target few-shot dataset with a peer dataset to create a data-unbalanced conditional generation. Our approach includes a class embedding method that separates the class space from the latent space, and we use a direction loss based on pre-trained CLIP to improve image diversity. Experiments on various few-shot datasets demonstrate the advancement of the proposed PIP, especially reduces the training requirements of few-shot image generation.

Fei Wang, Li Shen, Liang Ding et al.

Zeroth-Order optimization presents a promising memory-efficient paradigm for fine-tuning Large Language Models by relying solely on forward passes. However, its practical adoption is severely constrained by slow wall-clock convergence and high estimation variance. In this work, we dissect the runtime characteristics of ZO algorithms and identify a critical system bottleneck where the generation of perturbations and parameter updates accounts for over 40% of the training latency. We argue that the standard uniform exploration strategy is fundamentally flawed as it fails to account for the heterogeneous sensitivity of layers in deep networks, resulting in computationally wasteful blind searches. To address this structural mismatch, we propose AdaLeZO, an Adaptive Layer-wise ZO optimization framework. By formulating the layer selection process as a non-stationary Multi-Armed Bandit problem, AdaLeZO dynamically allocates the limited perturbation budget to the most sensitive parameters. We further introduce an Inverse Probability Weighting mechanism based on sampling with replacement, which guarantees unbiased gradient estimation while effectively acting as a temporal denoiser to reduce variance. Extensive experiments on LLaMA and OPT models ranging from 6.7B to 30B parameters demonstrate that AdaLeZO achieves 1.7x to 3.0x wall-clock acceleration compared to state-of-the-art methods. Crucially, AdaLeZO functions as a universal plug-and-play module that seamlessly enhances the efficiency of existing ZO optimizers without incurring additional memory overhead.

Rui Xia, Chao Xue, Boyu Deng et al.

Current researches indicate that inductive bias (IB) can improve Vision Transformer (ViT) performance. However, they introduce a pyramid structure concurrently to counteract the incremental FLOPs and parameters caused by introducing IB. This structure destroys the unification of computer vision and natural language processing (NLP) and complicates the model. We study an NLP model called LSRA, which introduces IB with a pyramid-free structure. We analyze why it outperforms ViT, discovering that introducing IB increases the share of high-frequency data in each layer, giving "attention" to more information. As a result, the heads notice more diverse information, showing better performance. To further explore the potential of transformers, we propose EIT, which Efficiently introduces IB to ViT with a novel decreasing convolutional structure under a pyramid-free structure. EIT achieves competitive performance with the state-of-the-art (SOTA) methods on ImageNet-1K and achieves SOTA performance over the same scale models which have the pyramid-free structure.

Chao Xue, Yao Wang, Mengqiao Liu et al.

Recent advancements in the Generative Reward Model (GRM) have demonstrated its potential to enhance the reasoning abilities of LLMs through Chain-of-Thought (CoT) prompting. Despite these gains, existing implementations of GRM suffer from two critical limitations. First, CoT prompting is applied indiscriminately to all inputs regardless of their inherent complexity. This introduces unnecessary computational costs for tasks amenable to fast, direct inference. Second, existing approaches primarily rely on voting-based mechanisms to evaluate CoT outputs, which often lack granularity and precision in assessing reasoning quality. In this paper, we propose E-GRM, an efficient generative reward modeling framework grounded in model-internal uncertainty. E-GRM leverages the convergence behavior of parallel model generations to estimate uncertainty and selectively trigger CoT reasoning only when needed, without relying on handcrafted features or task-dependent signals. To improve reward fidelity, we introduce a lightweight discriminative scorer trained with a hybrid regression--ranking objective to provide fine-grained evaluation of reasoning paths. Experiments on multiple reasoning benchmarks show that E-GRM substantially reduces inference cost while consistently improving answer accuracy, demonstrating that model-internal uncertainty is an effective and general signal for efficient reasoning-aware reward modeling.

Chao Xue, Yao Wang, Mengqiao Liu et al.

Supervised Fine-Tuning (SFT) is the standard approach for adapting large language models (LLMs) to downstream tasks. However, we observe a persistent failure mode: even after convergence, models often fail to correctly reproduce a subset of their own supervised training data. We refer to this behavior as the Incomplete Learning Phenomenon(ILP). This paper presents the first systematic study of ILP in LLM fine-tuning. We formalize ILP as post-training failure to internalize supervised instances and demonstrate its prevalence across multiple model families, domains, and datasets. Through controlled analyses, we identify five recurrent sources of incomplete learning: (1) missing prerequisite knowledge in the pre-trained model, (2) conflicts between SFT supervision and pre-training knowledge, (3) internal inconsistencies within SFT data, (4) left-side forgetting during sequential fine-tuning, and (5) insufficient optimization for rare or complex patterns. We introduce a diagnostic-first framework that maps unlearned samples to these causes using observable training and inference signals, and study several targeted mitigation strategies as causal interventions. Experiments on Qwen, LLaMA, and OLMo2 show that incomplete learning is widespread and heterogeneous, and that improvements in aggregate metrics can mask persistent unlearned subsets. The findings highlight the need for fine-grained diagnosis of what supervised fine-tuning fails to learn, and why.

Zekai Lin, Chao Xue, Di Liang et al.

Supervised Fine-Tuning (SFT) of large language models often suffers from task interference and catastrophic forgetting. Recent approaches alleviate this issue by isolating task-critical parameters during training. However, these methods represent a static solution to a dynamic problem, assuming that parameter importance remains fixed once identified. In this work, we empirically demonstrate that parameter importance exhibits temporal drift over the course of training. To address this, we propose Evolving Parameter Isolation (EPI), a fine-tuning framework that adapts isolation decisions based on online estimates of parameter importance. Instead of freezing a fixed subset of parameters, EPI periodically updates isolation masks using gradient-based signals, enabling the model to protect emerging task-critical parameters while releasing outdated ones to recover plasticity. Experiments on diverse multi-task benchmarks demonstrate that EPI consistently reduces interference and forgetting compared to static isolation and standard fine-tuning, while improving overall generalization. Our analysis highlights the necessity of synchronizing isolation mechanisms with the evolving dynamics of learning diverse abilities.

Wentao Tan, Qiong Cao, Chao Xue et al.

The chart-to-code generation task requires MLLMs to convert chart images into executable code. This task faces two main challenges: limited data diversity and the difficulty of maintaining visual consistency between generated charts and the original ones. Existing datasets mainly rely on synthetic seed data to prompt GPT models for code generation, resulting in homogeneous samples that limit model generalization to real-world chart styles. To address this, we propose ReChartPrompt, leveraging real-world, human-designed charts extracted from arXiv papers as prompts. By harnessing the rich content and diverse visual styles of arXiv charts, we construct ReChartPrompt-240K, a large-scale and highly diverse dataset that better reflects realistic chart variations. For the second challenge, although SFT improves code understanding by optimizing next-token prediction, it does not provide direct supervision on visual features. As a result, it often fails to guarantee that the generated charts visually match the original ones. To address this, we propose ChartSimRL, a GRPO-based reinforcement learning algorithm guided by a novel chart similarity reward. This reward consists of two components: attribute similarity, which measures the overlap of chart attributes like layout and color between the generated and original charts, and visual similarity, which evaluates overall visual features, including texture, using convolutional neural networks. Unlike traditional text-based rewards, our reward accounts for the multimodal nature of the chart-to-code generation task, significantly enhancing the model's ability to accurately reproduce charts. Integrating ReChartPrompt and ChartSimRL, we develop the ChartMaster model, achieving SOTA results among 7B-parameter models and rivaling GPT-4o on various chart-to-code benchmarks. All resources are available at https://github.com/WentaoTan/ChartMaster.

Wentao Tan, Qiong Cao, Yibing Zhan et al.

Achieving human-like reasoning capabilities in Multimodal Large Language Models (MLLMs) has long been a goal. Current methods primarily focus on synthesizing positive rationales, typically relying on manual annotations or complex systems. Moreover, they often overlook negative reasoning, which limits the model's generalization ability and robustness in multimodal inference. To address this gap, we propose a novel framework: \textbf{S}elf-Aligning \textbf{M}ultimodal Reasoning with \textbf{A}nswer-O\textbf{r}iented Chain-of-\textbf{T}hought (SMART). SMART employs an answer-oriented chain-of-thought (AoT) prompt to automatically construct high-quality data. Drawing inspiration from human proof-based strategies, AoT leverages both correct and incorrect answers to extract key visual information that links questions and answers. When provided with correct answers, the model produces strong positive rationales. Conversely, when correct answers are replaced with incorrect alternatives, the model generates an erroneous yet compelling reasoning path, serving as a form of discriminative negative rationale. Models trained with AoT-generated data outperform those trained on manually annotated datasets, demonstrating superior reasoning capabilities. Consequently, SMART establishes an iterative generation-optimization method that continually enhances the model's reasoning skills. Experiments indicate that the SMART framework significantly improves various MLLMs, regardless of model architecture, parameter size, or pre-training dataset. The code is available at https://github.com/WentaoTan/SMART.

Fei Wang, Li Shen, Liang Ding et al.

Large Language Models excel at natural language processing tasks, but their massive size leads to high computational and storage demands. Recent works have sought to reduce their model size through layer-wise structured pruning. However, they tend to ignore retaining the capabilities in the pruned part. In this work, we re-examine structured pruning paradigms and uncover several key limitations: 1) notable performance degradation due to direct layer removal, 2) incompetent linear weight layer aggregation, and 3) the lack of effective post-training recovery mechanisms. To address these limitations, we propose CoMe, including a progressive layer pruning framework with a Concatenation-based Merging technology and a hierarchical distillation post-training process. Specifically, we introduce a channel sensitivity metric that utilizes activation intensity and weight norms for fine-grained channel selection. Subsequently, we employ a concatenation-based layer merging method to fuse the most critical channels across adjacent layers, enabling progressive model size reduction. Finally, we propose a hierarchical distillation protocol that leverages the correspondences between the original and pruned model layers established during pruning, thereby enabling efficient knowledge transfer. Experiments on seven benchmarks show that CoMe achieves state-of-the-art performance; when pruning 30% of LLaMA-2-7b's parameters, the pruned model retains 83% of its original average accuracy. Our code is available at https://github.com/MPI-Lab/CoMe.

Zilong Zhang, Chujie Qin, Chunle Guo et al.

This work presents Robust Representation Learning via Adaptive Mask (RAM++), a two-stage framework for all-in-one image restoration. RAM++ integrates high-level semantic understanding with low-level texture generation to achieve content-oriented robust restoration. It addresses the limitations of existing degradation-oriented methods in extreme scenarios (e.g., degradations strongly coupled with image structures). RAM++ also mitigates common challenges such as unbalanced performance across tasks, overfitting to seen degradations, and weak generalization to unseen ones through three key designs: 1) Adaptive Semantic-Aware Mask (AdaSAM): a pretraining strategy that applies pixel-level masks to semantically rich and textured regions. This design enables the network to learn both generative priors and image content priors from various degradations. 2) Mask Attribute Conductance (MAC): a selective fine-tuning strategy that adjusts the layers with higher contributions to bridge the integrity gap between masked pretraining and full-image fine-tuning while retaining learned priors. 3) Robust Feature Regularization (RFR): a strategy that leverages DINOv2's semantically consistent and degradation-invariant representations, together with efficient feature fusion, to achieve faithful and semantically coherent restoration. With these designs, RAM++ achieves robust, well-balanced, and state-of-the-art performance across seen, unseen, extreme, and mixed degradations. Our code and model will be released at https://github.com/DragonisCV/RAM

Dan Song, Shumeng Huo, Wenhui Li et al.

The classification and recognition of maritime objects are crucial for enhancing maritime safety, monitoring, and intelligent sea environment prediction. However, existing unsupervised methods for maritime object classification often struggle with the long-tail data distributions in both object categories and weather conditions. In this paper, we construct a dataset named AIMO produced by large-scale generative models with diverse weather conditions and balanced object categories, and collect a dataset named RMO with real-world images where long-tail issue exists. We propose a novel domain adaptation approach that leverages AIMO (source domain) to address the problem of limited labeled data, unbalanced distribution and domain shift in RMO (target domain), enhance the generalization of source features with the Vision-Language Models such as CLIP, and propose a difficulty score for curriculum learning to optimize training process. Experimental results shows that the proposed method significantly improves the classification accuracy, particularly for samples within rare object categories and weather conditions. Datasets and codes will be publicly available at https://github.com/honoria0204/AIMO.

Canhui Wu, Qiong Cao, Chao Xue et al.

Recent advances in large language models (LLMs) have leveraged explicit Chain-of-Thought (CoT) prompting to improve reasoning accuracy. However, most existing methods primarily compress verbose reasoning outputs. These Long-to-Short transformations aim to improve efficiency, but still rely on explicit reasoning during inference. In this work, we introduce \textbf{3TF} (\textbf{T}hought-\textbf{T}raining and \textbf{T}hought-\textbf{F}ree inference), a framework for efficient reasoning that takes a Short-to-Long perspective. We first train a hybrid model that can operate in both reasoning and non-reasoning modes, and then further train it on CoT-annotated data to internalize structured reasoning, while enforcing concise, thought-free outputs at inference time using the no-reasoning mode. Unlike compression-based approaches, 3TF improves the reasoning quality of non-reasoning outputs, enabling models to perform rich internal reasoning implicitly while keeping external outputs short. Empirically, 3TF-trained models obtain large improvements on reasoning benchmarks under thought-free inference, demonstrating that high quality reasoning can be learned and executed implicitly without explicit step-by-step generation.

Chao Xue, Di Liang, Pengfei Wang et al.

Many models that leverage knowledge graphs (KGs) have recently demonstrated remarkable success in question answering (QA) tasks. In the real world, many facts contained in KGs are time-constrained thus temporal KGQA has received increasing attention. Despite the fruitful efforts of previous models in temporal KGQA, they still have several limitations. (I) They adopt pre-trained language models (PLMs) to obtain question representations, while PLMs tend to focus on entity information and ignore entity transfer caused by temporal constraints, and finally fail to learn specific temporal representations of entities. (II) They neither emphasize the graph structure between entities nor explicitly model the multi-hop relationship in the graph, which will make it difficult to solve complex multi-hop question answering. To alleviate this problem, we propose a novel Question Calibration and Multi-Hop Modeling (QC-MHM) approach. Specifically, We first calibrate the question representation by fusing the question and the time-constrained concepts in KG. Then, we construct the GNN layer to complete multi-hop message passing. Finally, the question representation is combined with the embedding output by the GNN to generate the final prediction. Empirical results verify that the proposed model achieves better performance than the state-of-the-art models in the benchmark dataset. Notably, the Hits@1 and Hits@10 results of QC-MHM on the CronQuestions dataset's complex questions are absolutely improved by 5.1% and 1.2% compared to the best-performing baseline. Moreover, QC-MHM can generate interpretable and trustworthy predictions.

Wentao Tan, Qiong Cao, Yibing Zhan et al.

Human preference alignment can greatly enhance Multimodal Large Language Models (MLLMs), but collecting high-quality preference data is costly. A promising solution is the self-evolution strategy, where models are iteratively trained on data they generate. However, current techniques still rely on human- or GPT-annotated data and sometimes require additional models or ground truth answers. To address these issues, we propose a novel multimodal self-evolution framework that enables the model to autonomously generate high-quality questions and answers using only unannotated images. First, we implement an image-driven self-questioning mechanism, allowing the model to create and evaluate questions based on image content, regenerating them if they are irrelevant or unanswerable. This sets a strong foundation for answer generation. Second, we introduce an answer self-enhancement technique, starting with image captioning to improve answer quality. We also use corrupted images to generate rejected answers, forming distinct preference pairs for optimization. Finally, we incorporate an image content alignment loss function alongside Direct Preference Optimization (DPO) loss to reduce hallucinations, ensuring the model focuses on image content. Experiments show that our framework performs competitively with methods using external information, offering a more efficient and scalable approach to MLLMs.

Fei Wang, Li Shen, Liang Ding et al.

Fine-tuning is powerful for adapting large language models to downstream tasks, but it often results in huge memory usages. A promising approach to mitigate this is using Zeroth-Order (ZO) optimization, which estimates gradients to replace First-Order (FO) gradient calculations, albeit with longer training time due to its stochastic nature. By revisiting the Memory-efficient ZO (MeZO) optimizer, we discover that the full-parameter perturbation and updating processes consume over 50% of its overall fine-tuning time cost. Based on these observations, we introduce a novel layer-wise sparse computation and memory efficient ZO optimizer, named LeZO. LeZO treats layers as fundamental units for sparsification and dynamically perturbs different parameter subsets in each step to achieve full-parameter fine-tuning. LeZO incorporates layer-wise parameter sparsity in the process of simultaneous perturbation stochastic approximation (SPSA) and ZO stochastic gradient descent (ZO-SGD). It achieves accelerated computation during perturbation and updating processes without additional memory overhead. We conduct extensive experiments with the OPT model family on the SuperGLUE benchmark and two generative tasks. The experiments show that LeZO accelerates training without compromising the performance of ZO optimization. Specifically, it achieves over 3x speedup compared to MeZO on the SST-2, BoolQ, and Copa tasks.

Chao Xue, Ziyuan Gao

Text semantic matching requires nuanced understanding of both structural relationships and fine-grained semantic distinctions. While pre-trained language models excel at capturing token-level interactions, they often overlook hierarchical structural patterns and struggle with subtle semantic discrimination. In this paper, we proposed StructCoh, a graph-enhanced contrastive learning framework that synergistically combines structural reasoning with representation space optimization. Our approach features two key innovations: (1) A dual-graph encoder constructs semantic graphs via dependency parsing and topic modeling, then employs graph isomorphism networks to propagate structural features across syntactic dependencies and cross-document concept nodes. (2) A hierarchical contrastive objective enforces consistency at multiple granularities: node-level contrastive regularization preserves core semantic units, while graph-aware contrastive learning aligns inter-document structural semantics through both explicit and implicit negative sampling strategies. Experiments on three legal document matching benchmarks and academic plagiarism detection datasets demonstrate significant improvements over state-of-the-art methods. Notably, StructCoh achieves 86.7% F1-score (+6.2% absolute gain) on legal statute matching by effectively identifying argument structure similarities.

Jiaxing Li, Wei Liu, Chao Xue et al.

Bayesian Optimization (BO) is a sample-efficient black-box optimizer commonly used in search spaces where hyperparameters are independent. However, in many practical AutoML scenarios, there will be dependencies among hyperparameters, forming a conditional search space, which can be partitioned into structurally distinct subspaces. The structure and dimensionality of hyperparameter configurations vary across these subspaces, challenging the application of BO. Some previous BO works have proposed solutions to develop multiple Gaussian Process models in these subspaces. However, these approaches tend to be inefficient as they require a substantial number of observations to guarantee each GP's performance and cannot capture relationships between hyperparameters across different subspaces. To address these issues, this paper proposes a novel approach to model the response surfaces of all subspaces in one, which can model the relationships between hyperparameters elegantly via a self-attention mechanism. Concretely, we design a structure-aware hyperparameter embedding to preserve the structural information. Then, we introduce an attention-based deep feature extractor, capable of projecting configurations with different structures from various subspaces into a unified feature space, where the response surfaces can be formulated using a single standard Gaussian Process. The empirical results on a simulation function, various real-world tasks, and HPO-B benchmark demonstrate that our proposed approach improves the efficacy and efficiency of BO within conditional search spaces.

Xiaoxing Wang, Jiaxing Li, Chao Xue et al.

BayesianOptimization(BO) is a sample-efficient black-box optimizer, and extensive methods have been proposed to build the absolute function response of the black-box function through a probabilistic surrogate model, including Tree-structured Parzen Estimator (TPE), random forest (SMAC), and Gaussian process (GP). However, few methods have been explored to estimate the relative rankings of candidates, which can be more robust to noise and have better practicality than absolute function responses, especially when the function responses are intractable but preferences can be acquired. To this end, we propose a novel ranking-based surrogate model based on the Poisson process and introduce an efficient BO framework, namely Poisson Process Bayesian Optimization (PoPBO). Two tailored acquisition functions are further derived from classic LCB and EI to accommodate it. Compared to the classic GP-BO method, our PoPBO has lower computation costs and better robustness to noise, which is verified by abundant experiments. The results on both simulated and real-world benchmarks, including hyperparameter optimization (HPO) and neural architecture search (NAS), show the effectiveness of PoPBO.

Canhui Wu, Qiong Cao, Chang Li et al.

Large Reasoning Models (LRMs) demonstrate strong performance on complex tasks but often suffer from excessive verbosity, known as "overthinking." Existing solutions via reinforcement learning (RL) typically penalize generated tokens to promote conciseness. However, these methods encounter two challenges: responses with fewer tokens do not always correspond to fewer reasoning steps, and models may develop hacking behavior in later stages of training by discarding reasoning steps to minimize token usage. In this work, we introduce \textbf{Step Pruner (SP)}, an RL framework that steers LRMs toward more efficient reasoning by favoring compact reasoning steps. Our step-aware reward function prioritizes correctness while imposing penalties for redundant steps, and withholds rewards for incorrect responses to prevent the reinforcement of erroneous reasoning. Moreover, we propose a dynamic stopping mechanism: when the length of any output step exceeds the upper limit, we halt updates to prevent hacking behavior caused by merging steps. Extensive experiments across four reasoning benchmarks demonstrate that SP achieves state-of-the-art accuracy while significantly reducing response length. For instance, on AIME24, SP reduces token usage by \textbf{69.7\%}.

Zhiting Gao, Dan Song, Diqiong Jiang et al.

Motion generation is essential for animating virtual characters and embodied agents. While recent text-driven methods have made significant strides, they often struggle with achieving precise alignment between linguistic descriptions and motion semantics, as well as with the inefficiencies of slow, multi-step inference. To address these issues, we introduce TMR++ Aligned Preference Optimization (TAPO), an innovative framework that aligns subtle motion variations with textual modifiers and incorporates iterative adjustments to reinforce semantic grounding. To further enable real-time synthesis, we propose MotionFLUX, a high-speed generation framework based on deterministic rectified flow matching. Unlike traditional diffusion models, which require hundreds of denoising steps, MotionFLUX constructs optimal transport paths between noise distributions and motion spaces, facilitating real-time synthesis. The linearized probability paths reduce the need for multi-step sampling typical of sequential methods, significantly accelerating inference time without sacrificing motion quality. Experimental results demonstrate that, together, TAPO and MotionFLUX form a unified system that outperforms state-of-the-art approaches in both semantic consistency and motion quality, while also accelerating generation speed. The code and pretrained models will be released.

Chang Li, Yaren Zhang, Haoran Lv et al.

Large Language Models (LLMs) have shown remarkable reasoning ability through explicit Chain-of-Thought (CoT) prompting, but generating these step-by-step textual explanations is computationally expensive and slow. To overcome this, we aim to develop a framework for efficient, implicit reasoning, where the model "thinks" in a latent space without generating explicit text for every step. We propose that these latent thoughts can be modeled as temporally-extended abstract actions, or options, within a hierarchical reinforcement learning framework. To effectively learn a diverse library of options as latent embeddings, we first introduce the Variational Markovian Option Critic (VMOC), an off-policy algorithm that uses variational inference within the HiT-MDP framework. To provide a rigorous foundation for using these options as an abstract reasoning space, we extend the theory of continuous MDP homomorphisms. This proves that learning a policy in the simplified, abstract latent space, for which VMOC is suited, preserves the optimality of the solution to the original, complex problem. Finally, we propose a cold-start procedure that leverages supervised fine-tuning (SFT) data to distill human reasoning demonstrations into this latent option space, providing a rich initialization for the model's reasoning capabilities. Extensive experiments demonstrate that our approach achieves strong performance on complex logical reasoning benchmarks and challenging locomotion tasks, validating our framework as a principled method for learning abstract skills for both language and control.

Mengting Hu, Shiwan Zhao, Honglei Guo et al.

Aspect category detection (ACD) in sentiment analysis aims to identify the aspect categories mentioned in a sentence. In this paper, we formulate ACD in the few-shot learning scenario. However, existing few-shot learning approaches mainly focus on single-label predictions. These methods can not work well for the ACD task since a sentence may contain multiple aspect categories. Therefore, we propose a multi-label few-shot learning method based on the prototypical network. To alleviate the noise, we design two effective attention mechanisms. The support-set attention aims to extract better prototypes by removing irrelevant aspects. The query-set attention computes multiple prototype-specific representations for each query instance, which are then used to compute accurate distances with the corresponding prototypes. To achieve multi-label inference, we further learn a dynamic threshold per instance by a policy network. Extensive experimental results on three datasets demonstrate that the proposed method significantly outperforms strong baselines.

Tao Wang, Junsong Wang, Chang Xu et al.

Model quantization is a widely used technique to compress and accelerate deep neural network (DNN) inference, especially when deploying to edge or IoT devices with limited computation capacity and power consumption budget. The uniform bit width quantization across all the layers is usually sub-optimal and the exploration of hybrid quantization for different layers is vital for efficient deep compression. In this paper, we employ the meta learning method to automatically realize low-bit hybrid quantization of neural networks. A MetaQuantNet, together with a Quantization function, are trained to generate the quantized weights for the target DNN. Then, we apply a genetic algorithm to search the best hybrid quantization policy that meets compression constraints. With the best searched quantization policy, we subsequently retrain or finetune to further improve the performance of the quantized target network. Extensive experiments demonstrate the performance of searched hybrid quantization scheme surpass that of uniform bitwidth counterpart. Compared to the existing reinforcement learning (RL) based hybrid quantization search approach that relies on tedious explorations, our meta learning approach is more efficient and effective for any compression requirements since the MetaQuantNet only needs be trained once.

Sivan Doveh, Eli Schwartz, Chao Xue et al.

Few-Shot Learning (FSL) is a topic of rapidly growing interest. Typically, in FSL a model is trained on a dataset consisting of many small tasks (meta-tasks) and learns to adapt to novel tasks that it will encounter during test time. This is also referred to as meta-learning. Another topic closely related to meta-learning with a lot of interest in the community is Neural Architecture Search (NAS), automatically finding optimal architecture instead of engineering it manually. In this work, we combine these two aspects of meta-learning. So far, meta-learning FSL methods have focused on optimizing parameters of pre-defined network architectures, in order to make them easily adaptable to novel tasks. Moreover, it was observed that, in general, larger architectures perform better than smaller ones up to a certain saturation point (where they start to degrade due to over-fitting). However, little attention has been given to explicitly optimizing the architectures for FSL, nor to an adaptation of the architecture at test time to particular novel tasks. In this work, we propose to employ tools inspired by the Differentiable Neural Architecture Search (D-NAS) literature in order to optimize the architecture for FSL without over-fitting. Additionally, to make the architecture task adaptive, we propose the concept of `MetAdapt Controller' modules. These modules are added to the model and are meta-trained to predict the optimal network connections for a given novel task. Using the proposed approach we observe state-of-the-art results on two popular few-shot benchmarks: miniImageNet and FC100.

Atin Sood, Benjamin Elder, Benjamin Herta et al.

Application of neural networks to a vast variety of practical applications is transforming the way AI is applied in practice. Pre-trained neural network models available through APIs or capability to custom train pre-built neural network architectures with customer data has made the consumption of AI by developers much simpler and resulted in broad adoption of these complex AI models. While prebuilt network models exist for certain scenarios, to try and meet the constraints that are unique to each application, AI teams need to think about developing custom neural network architectures that can meet the tradeoff between accuracy and memory footprint to achieve the tight constraints of their unique use-cases. However, only a small proportion of data science teams have the skills and experience needed to create a neural network from scratch, and the demand far exceeds the supply. In this paper, we present NeuNetS : An automated Neural Network Synthesis engine for custom neural network design that is available as part of IBM's AI OpenScale's product. NeuNetS is available for both Text and Image domains and can build neural networks for specific tasks in a fraction of the time it takes today with human effort, and with accuracy similar to that of human-designed AI models.