Yuwei Hu

Sai Zhang, Yuwei Hu, Yuchuan Wu et al.

A slot value might be provided segment by segment over multiple-turn interactions in a dialog, especially for some important information such as phone numbers and names. It is a common phenomenon in daily life, but little attention has been paid to it in previous work. To fill the gap, this paper defines a new task named Sub-Slot based Task-Oriented Dialog (SSTOD) and builds a Chinese dialog dataset SSD for boosting research on SSTOD. The dataset includes a total of 40K dialogs and 500K utterances from four different domains: Chinese names, phone numbers, ID numbers and license plate numbers. The data is well annotated with sub-slot values, slot values, dialog states and actions. We find some new linguistic phenomena and interactive manners in SSTOD which raise critical challenges of building dialog agents for the task. We test three state-of-the-art dialog models on SSTOD and find they cannot handle the task well on any of the four domains. We also investigate an improved model by involving slot knowledge in a plug-in manner. More work should be done to meet the new challenges raised from SSTOD which widely exists in real-life applications. The dataset and code are publicly available via https://github.com/shunjiu/SSTOD.

Yuwei Hu, Jiajie Li, Zhongming Yu et al.

Graph neural networks (GNNs), which have emerged as an effective method for handling machine learning tasks on graphs, bring a new approach to building recommender systems, where the task of recommendation can be formulated as the link prediction problem on user-item bipartite graphs. Training GNN-based recommender systems (GNNRecSys) on large graphs incurs a large memory footprint, easily exceeding the DRAM capacity on a typical server. Existing solutions resort to distributed subgraph training, which is inefficient due to the high cost of dynamically constructing subgraphs and significant redundancy across subgraphs. The emerging persistent memory technologies provide a significantly larger memory capacity than DRAMs at an affordable cost, making single-machine GNNRecSys training feasible, which eliminates the inefficiencies in distributed training. One major concern of using persistent memory devices for GNNRecSys is their relatively low bandwidth compared with DRAMs. This limitation can be particularly detrimental to achieving high performance for GNNRecSys workloads since their dominant compute kernels are sparse and memory access intensive. To understand whether persistent memory is a good fit for GNNRecSys training, we perform an in-depth characterization of GNNRecSys workloads and a comprehensive analysis of their performance on a persistent memory device, namely, Intel Optane. Based on the analysis, we provide guidance on how to configure Optane for GNNRecSys workloads. Furthermore, we present techniques for large-batch training to fully realize the advantages of single-machine GNNRecSys training. Our experiment results show that with the tuned batch size and optimal system configuration, Optane-based single-machine GNNRecSys training outperforms distributed training by a large margin, especially when handling deep GNN models.

Shiyin Lu, Yang Li, Yu Xia et al.

We present Ovis2.5, a successor to Ovis2 designed for native-resolution visual perception and strong multimodal reasoning. Ovis2.5 integrates a native-resolution vision transformer that processes images at their native, variable resolutions, avoiding the degradation from fixed-resolution tiling and preserving both fine detail and global layout -- crucial for visually dense content like complex charts. To strengthen reasoning, we train the model to move beyond linear chain-of-thought and perform reflection -- including self-checking and revision. This advanced capability is exposed as an optional "thinking mode" at inference time, allowing users to trade latency for enhanced accuracy on difficult inputs. The model is trained via a comprehensive five-phase curriculum that progressively builds its skills. The process begins with foundational visual and multimodal pretraining, advances through large-scale instruction tuning, and culminates in alignment and reasoning enhancement using DPO and GRPO. To scale these upgrades efficiently, we employ multimodal data packing and hybrid parallelism, yielding a significant end-to-end speedup. We release two open-source models: Ovis2.5-9B and Ovis2.5-2B. The latter continues the "small model, big performance" philosophy of Ovis2, making it ideal for resource-constrained, on-device scenarios. On the OpenCompass multimodal leaderboard, Ovis2.5-9B averages 78.3, marking a substantial improvement over its predecessor, Ovis2-8B, and achieving state-of-the-art results among open-source MLLMs in the sub-40B parameter range; Ovis2.5-2B scores 73.9, establishing SOTA for its size. Beyond aggregate scores, Ovis2.5 achieves leading results on STEM benchmarks, exhibits strong capabilities on grounding and video tasks, and achieves open-source SOTA at its scale for complex chart analysis.

Jinhu Qi, Shuai Yan, Yibo Zhang et al.

With the development of the modern social economy, tourism has become an important way to meet people's spiritual needs, bringing development opportunities to the tourism industry. However, existing large language models (LLMs) face challenges in personalized recommendation capabilities and the generation of content that can sometimes produce hallucinations. This study proposes an optimization scheme for Tibet tourism LLMs based on retrieval-augmented generation (RAG) technology. By constructing a database of tourist viewpoints and processing the data using vectorization techniques, we have significantly improved retrieval accuracy. The application of RAG technology effectively addresses the hallucination problem in content generation. The optimized model shows significant improvements in fluency, accuracy, and relevance of content generation. This research demonstrates the potential of RAG technology in the standardization of cultural tourism information and data analysis, providing theoretical and technical support for the development of intelligent cultural tourism service systems.

Yuwei Hu, Bryce L. Ferguson

In many engineered systems, agents make decisions under incomplete information, creating opportunities for a planner to influence decentralized behavior through signaling. We study how such signaling can be designed in parallel-network, affine latency congestion games when users may not interpret recommendations using the same beliefs assumed by the planner. To do so, we consider Bayesian congestion games with private recommendations and formulate a robust information design problem in which obedience must hold uniformly over a neighborhood of a nominal prior. This addresses the previously uncharacterized issue of whether obedience itself remains reliable under belief heterogeneity, rather than only under the single prior used at the design stage. We characterize policy-level robustness radii, identify regimes in which the robust obedience region remains nonempty, and analyze the resulting robustness--performance tradeoff through a robust value function whose optimal cost is monotone in the robustness requirement and whose local sensitivity is governed by the active obedience constraints.

Jiaqi Tang, Yu Xia, Yi-Feng Wu et al.

The advent of autonomous agents is transforming interactions with Graphical User Interfaces (GUIs) by employing natural language as a powerful intermediary. Despite the predominance of Supervised Fine-Tuning (SFT) methods in current GUI agents for achieving spatial localization, these methods face substantial challenges due to their limited capacity to accurately perceive positional data. Existing strategies, such as reinforcement learning, often fail to assess positional accuracy effectively, thereby restricting their utility. In response, we introduce Location Preference Optimization (LPO), a novel approach that leverages locational data to optimize interaction preferences. LPO uses information entropy to predict interaction positions by focusing on zones rich in information. Besides, it further introduces a dynamic location reward function based on physical distance, reflecting the varying importance of interaction positions. Supported by Group Relative Preference Optimization (GRPO), LPO facilitates an extensive exploration of GUI environments and significantly enhances interaction precision. Comprehensive experiments demonstrate LPO's superior performance, achieving SOTA results across both offline benchmarks and real-world online evaluations. Our code will be made publicly available soon, at https://github.com/AIDC-AI/LPO.

Yanqi Zhang, Yuwei Hu, Runyuan Zhao et al.

Large language models (LLMs) demonstrate remarkable capabilities but face substantial serving costs due to their high memory demands, with the key-value (KV) cache being a primary bottleneck. State-of-the-art KV cache compression techniques, such as quantization and pruning, apply uniform treatment to both keys and values, and discard unimportant tokens entirely, overlooking the fine-grained distinctions in the significance of individual KV cache components. To address such limitations, we introduce \textit{DiffKV}, a novel framework for efficient KV cache compression that exploits three levels of differentiation in the KV cache: (1) the differing impact of keys and values on attention computation, (2) the varying importance of tokens, and (3) the diverse dynamic sparsity patterns across attention heads. These levels of differentiation introduce irregular memory usage patterns across different requests and attention heads, posing significant scalability challenges for memory management. To address these challenges, DiffKV proposes an on-GPU memory manager that compacts fragmented free memory list into contiguous regions in parallel, effectively translating sparsity in the KV cache into performance gains. We evaluate DiffKV on several mainstream LLMs, including the emerging thinking models that generate extended chains of thought. DiffKV is able to compress the KV cache by $2.7\times$ to $5.7\times$ with near-lossless accuracy on complex workloads requiring sophisticated reasoning and long-generation capabilities, and enhances throughput by $1.9\times$ to $5.4\times$. Source codes of DiffKV are available at https://github.com/zyqCSL/DiffKV.

Ziyu Zhao, Leilei Gan, Guoyin Wang et al.

Low-Rank Adaptation (LoRA) offers an efficient way to fine-tune large language models (LLMs). Its modular and plug-and-play nature allows the integration of various domain-specific LoRAs, enhancing LLM capabilities. Open-source platforms like Huggingface and Modelscope have introduced a new computational paradigm, Uploadable Machine Learning (UML). In UML, contributors use decentralized data to train specialized adapters, which are then uploaded to a central platform to improve LLMs. This platform uses these domain-specific adapters to handle mixed-task requests requiring personalized service. Previous research on LoRA composition either focuses on specific tasks or fixes the LoRA selection during training. However, in UML, the pool of LoRAs is dynamically updated with new uploads, requiring a generalizable selection mechanism for unseen LoRAs. Additionally, the mixed-task nature of downstream requests necessitates personalized services. To address these challenges, we propose Retrieval-Augmented Mixture of LoRA Experts (RAMoLE), a framework that adaptively retrieves and composes multiple LoRAs based on input prompts. RAMoLE has three main components: LoraRetriever for identifying and retrieving relevant LoRAs, an on-the-fly MoLE mechanism for coordinating the retrieved LoRAs, and efficient batch inference for handling heterogeneous requests. Experimental results show that RAMoLE consistently outperforms baselines, highlighting its effectiveness and scalability.

Sai Zhang, Yuwei Hu, Xiaojie Wang et al.

Reinforcement learning has been applied to train the dialog systems in many works. Previous approaches divide the dialog system into multiple modules including DST (dialog state tracking) and DP (dialog policy), and train these modules simultaneously. However, different modules influence each other during training. The errors from DST might misguide the dialog policy, and the system action brings extra difficulties for the DST module. To alleviate this problem, we propose Asynchronous Updating Reinforcement Learning framework (AURL) that updates the DST module and the DP module asynchronously under a cooperative setting. Furthermore, curriculum learning is implemented to address the problem of unbalanced data distribution during reinforcement learning sampling, and multiple user models are introduced to increase the dialog diversity. Results on the public SSD-PHONE dataset show that our method achieves a compelling result with a 31.37% improvement on the dialog success rate. The code is publicly available via https://github.com/shunjiu/AURL.

Tianqi Chen, Thierry Moreau, Ziheng Jiang et al.

There is an increasing need to bring machine learning to a wide diversity of hardware devices. Current frameworks rely on vendor-specific operator libraries and optimize for a narrow range of server-class GPUs. Deploying workloads to new platforms -- such as mobile phones, embedded devices, and accelerators (e.g., FPGAs, ASICs) -- requires significant manual effort. We propose TVM, a compiler that exposes graph-level and operator-level optimizations to provide performance portability to deep learning workloads across diverse hardware back-ends. TVM solves optimization challenges specific to deep learning, such as high-level operator fusion, mapping to arbitrary hardware primitives, and memory latency hiding. It also automates optimization of low-level programs to hardware characteristics by employing a novel, learning-based cost modeling method for rapid exploration of code optimizations. Experimental results show that TVM delivers performance across hardware back-ends that are competitive with state-of-the-art, hand-tuned libraries for low-power CPU, mobile GPU, and server-class GPUs. We also demonstrate TVM's ability to target new accelerator back-ends, such as the FPGA-based generic deep learning accelerator. The system is open sourced and in production use inside several major companies.

Yuwei Hu, Xinyi Huang, Zhewei Wei et al.

Large Language Models (LLMs) for Graph Reasoning have been extensively studied over the past two years, involving enabling LLMs to understand graph structures and reason on graphs to solve various graph problems, with graph algorithm problems being the most prevalent. Recent studies underscore the potential of LLMs in handling graph reasoning tasks, but their performance is underwhelming. In this work, we point out issues with existing methods and benchmarks, and rethink the direction that LLMs for graph reasoning should strive toward. We find that base models, e.g., GPT-4o-mini, are largely underestimated due to improper reasoning focus. Base models with reasoning focus redirected from replicating graph algorithms to designing them can easily solve most graph reasoning tasks in existing benchmarks. To truly evaluate the graph reasoning capabilities of LLMs, we construct a more challenging GraphAlgorithm benchmark, comprising 239 different graph problems and 3,041 test instances collected from 4 competition platforms. Finally, we introduce a simple and strong baseline Simple-Reasoning-Then-Coding (Simple-RTC)-which guides LLMs to design graph algorithms first and then code to address graph reasoning tasks. Simple-RTC achieves near-perfect accuracy on existing benchmarks and significantly outperforms GPT-4o-mini and all prior methods on the GraphAlgorithm benchmark. This strong baseline encourages further advancements in LLMs for Graph Reasoning in the future.

Mark Buckler, Neil Adit, Yuwei Hu et al.

Depthwise separable convolutions and frequency-domain convolutions are two recent ideas for building efficient convolutional neural networks. They are seemingly incompatible: the vast majority of operations in depthwise separable CNNs are in pointwise convolutional layers, but pointwise layers use 1x1 kernels, which do not benefit from frequency transformation. This paper unifies these two ideas by transforming the activations, not the kernels. Our key insights are that 1) pointwise convolutions commute with frequency transformation and thus can be computed in the frequency domain without modification, 2) each channel within a given layer has a different level of sensitivity to frequency domain pruning, and 3) each channel's sensitivity to frequency pruning is approximately monotonic with respect to frequency. We leverage this knowledge by proposing a new technique which wraps each pointwise layer in a discrete cosine transform (DCT) which is truncated to selectively prune coefficients above a given threshold as per the needs of each channel. To learn which frequencies should be pruned from which channels, we introduce a novel learned parameter which specifies each channel's pruning threshold. We add a new regularization term which incentivizes the model to decrease the number of retained frequencies while still maintaining task accuracy. Unlike weight pruning techniques which rely on sparse operators, our contiguous frequency band pruning results in fully dense computation. We apply our technique to MobileNetV2 and in the process reduce computation time by 22% and incur <1% accuracy degradation.

Yuwei Hu, Zihao Ye, Minjie Wang et al.

Graph neural networks (GNNs) are gaining increasing popularity as a promising approach to machine learning on graphs. Unlike traditional graph workloads where each vertex/edge is associated with a scalar, GNNs attach a feature tensor to each vertex/edge. This additional feature dimension, along with consequently more complex vertex- and edge-wise computations, has enormous implications on locality and parallelism, which existing graph processing systems fail to exploit. This paper proposes FeatGraph to accelerate GNN workloads by co-optimizing graph traversal and feature dimension computation. FeatGraph provides a flexible programming interface to express diverse GNN models by composing coarse-grained sparse templates with fine-grained user-defined functions (UDFs) on each vertex/edge. FeatGraph incorporates optimizations for graph traversal into the sparse templates and allows users to specify optimizations for UDFs with a feature dimension schedule (FDS). FeatGraph speeds up end-to-end GNN training and inference by up to 32x on CPU and 7x on GPU.

Ritchie Zhao, Yuwei Hu, Jordan Dotzel et al.

Quantization can improve the execution latency and energy efficiency of neural networks on both commodity GPUs and specialized accelerators. The majority of existing literature focuses on training quantized DNNs, while this work examines the less-studied topic of quantizing a floating-point model without (re)training. DNN weights and activations follow a bell-shaped distribution post-training, while practical hardware uses a linear quantization grid. This leads to challenges in dealing with outliers in the distribution. Prior work has addressed this by clipping the outliers or using specialized hardware. In this work, we propose outlier channel splitting (OCS), which duplicates channels containing outliers, then halves the channel values. The network remains functionally identical, but affected outliers are moved toward the center of the distribution. OCS requires no additional training and works on commodity hardware. Experimental evaluation on ImageNet classification and language modeling shows that OCS can outperform state-of-the-art clipping techniques with only minor overhead.

Ritchie Zhao, Yuwei Hu, Jordan Dotzel et al.

We propose unitary group convolutions (UGConvs), a building block for CNNs which compose a group convolution with unitary transforms in feature space to learn a richer set of representations than group convolution alone. UGConvs generalize two disparate ideas in CNN architecture, channel shuffling (i.e. ShuffleNet) and block-circulant networks (i.e. CirCNN), and provide unifying insights that lead to a deeper understanding of each technique. We experimentally demonstrate that dense unitary transforms can outperform channel shuffling in DNN accuracy. On the other hand, different dense transforms exhibit comparable accuracy performance. Based on these observations we propose HadaNet, a UGConv network using Hadamard transforms. HadaNets achieve similar accuracy to circulant networks with lower computation complexity, and better accuracy than ShuffleNets with the same number of parameters and floating-point multiplies.