arXiv subjects

Xiao Yan

Publications and source records attributed to Xiao Yan.

At least 19 recordsLinked to original sources

arXiv2026

Can Edge-Deployable Vision-Language Models Identify Species?

Camera traps often run in the field on edge hardware with limited or no connectivity, making small, locally-deployable vision-language models (VLMs) -- not frontier-scale ones -- the practically relevant class to evaluate for species identification. We test whether models in this deployment-relevant 2--8B range carry genuine taxonomic knowledge, evaluating four such VLMs (Qwen3-VL 2B/4B/8B, Gemma3 4B) against the domain-specific specialist BioCLIP (300M parameters) on a 96-species task, comparing clean iNaturalist photographs against camera-trap imagery from 6 LILA.science collections, on two independently-sampled evaluation sets. All models identify species far above chance, but every model -- general-purpose or specialist -- degrades sharply on field imagery (domain gaps of 9.6--26.6 percentage points, consistent across taxonomic levels and both evaluation sets), indicating the degradation reflects general image legibility rather than fine-grained discrimination failure. BioCLIP substantially outperforms every VLM tested (by 33.2--59.2 percentage points across an expanded 200-image sample for every model) despite its far smaller size, suggesting the gap reflects specialized training data rather than model scale; yet BioCLIP's own domain gap (18.0 points) is statistically indistinguishable from the best VLM's (22.3 points), suggesting the clean-to-field degradation itself is a property of the image-quality shift rather than a general-purpose-model weakness. Under open-set prompting, 5.9--9.6% of responses are syntactically valid but taxonomically nonexistent species names; the relative fabrication-rate ranking across models replicates exactly across both evaluation sets, a more robust finding than any single point estimate.

cs.AI↗

arXiv2026

CAM: Question Answering on Entity-Centric Videos with Continuous Extraction and Adaptive Querying

Memory facilitates question answering over long videos by extracting and retrieving facts to fit within the limited context windows of multimodal LLMs (MLLMs). Existing solutions typically extract independent memory entries from fixed-length video clips and thus cannot capture high-level semantics that need to be summarized over extended time periods, such as character traits and relations. Moreover, they rely solely on similarity-based retrieval and may fail to retrieve the fine-grained details required for question answering. To tackle these problems, we propose CAM, featuring continuous extraction for high-level semantics and adaptive querying for fine-grained details. In particular, CAM stores the entities and relations extracted from video clips in a knowledge graph. To capture the high-level semantics of each entity or relation, CAM summarizes the local subgraph of the target entity or relation once the subgraph reaches a predefined size. To retrieve the fine-grained details required for question answering, CAM supports multiple search methods, including knowledge graph traversal, video re-watching, and audio listening. It utilizes a planner-executor-verifier pipeline to adaptively compose these search methods according to question intent. Evaluations on three benchmarks show that CAM outperforms SOTA baselines and improves their accuracy by up to 23 percentage points. Code is available at https://github.com/Jake-Tian/CAM.

cs.CV↗

arXiv2026

Chimera: Efficient Multi-Vector Retrieval via GPU-CPU Co-Processing

Multi-vector retrieval has become an important primitive for fine-grained matching in information retrieval, with emerging applications in areas such as recommender systems and bioinformatics. However, its high computational complexity and memory costs make low-latency retrieval difficult. Prior systems have attempted to optimize query latency, but their designs remain CPU-centric. While GPUs offer substantial computational advantages, their limited memory capacity necessitates a heterogeneous architecture in which the dataset resides in host memory and the GPU serves as an accelerator. Existing GPU-based system, PLAID, is bottlenecked by CPU-GPU data movement, as vector data must be transferred from host memory to the GPU at query time. We propose Chimera, a GPU-CPU co-processing system for multi-vector retrieval that eliminates this transfer bottleneck. Chimera stores highly compressed, low-precision quantization codes on the GPU while maintaining high-precision data in CPU memory. At query time, it leverages GPU-resident data for efficient candidate generation and filtering, and further refines results through a GPU-CPU collaborative scoring scheme that completely avoids vector data transfer while enabling computation overlap. Experiments on real-world datasets demonstrate that Chimera significantly outperforms existing approaches, achieving up to 59.5x higher QPS at the same recall level.

cs.DB↗

arXiv2026

SiriusDeliver: Automating Data Warehouse Delivery at Tencent

Enterprise data warehouses (DWs) support business-critical analytics, but warehouse task delivery remains a complicated production process involving context retrieval, workflow configuration, code generation, platform submission, and failure diagnosis. Although large language models (LLMs) and coding agents have improved software development, they are insufficient for production DW delivery, which requires dependency-aware orchestration, lifecycle-aware artifact control, and continuous adaptation to evolving platform practices. We present SiriusDeliver, an end-to-end delivery automation agent for production warehouse task submission. SiriusDeliver integrates three components: a hierarchical delivery agent that orchestrates warehouse skills, an artifact lifecycle control module that verifies and revises artifacts before and after platform execution, and a trace-driven skill evolution mechanism that maintains reusable skills from delivery trajectories. We evaluate SiriusDeliver through offline datasets and large-scale production deployment on Tencent Cloud WeData. Offline experiments on real-world warehouse delivery cases show that SiriusDeliver improves delivery success and automation efficiency over representative baselines. During a two-month deployment across 6 business teams and 4 warehouse task types, SiriusDeliver served 3,600 monthly active users and supported 18,240 delivery sessions, achieving an 87.2% end-to-end success rate and a 73.5% autonomous submission rate. A one-month A/B test shows that SiriusDeliver reduces median delivery time from 228 to 23 minutes and engineer effort from 95 to 11 minutes, while maintaining comparable final delivery success.

cs.DB↗

arXiv2026

Metis: Bridging Text and Code Memory for Self-Evolving Agents

Self-evolving agents improve over time by distilling experience from past executions and reusing it in future tasks. Existing systems represent such experience either as natural-language text injected into the agent context or as code exposed as callable tools. However, the choice between these representations is typically made at design time rather than derived from the characteristics of the experience itself, leaving the trade-offs between them poorly understood. We present the first controlled study that isolates text memory and code memory over an identical set of experiences. Our results show that the two forms exhibit complementary trade-offs in construction cost, execution efficiency, and transferability, such that neither representation alone is sufficient. Guided by these findings, we propose Metis, a self-evolving agent system built on a hierarchical dual-representation memory. Metis organizes textual experience into execution plans, environment facts, and common pitfalls, and selectively crystallizes recurring plans into validated callable tools. This design combines the broad applicability of text memory with the execution efficiency of code memory while incurring tool-generation cost only when justified by repeated reuse. We evaluate Metis on AppWorld, a challenging benchmark for interactive agents. The results show that Metis improves task accuracy by up to 20.6% over ReAct while reducing execution cost by up to 22.8%. Compared with representative self-evolving agent systems, Metis consistently achieves a better balance between accuracy, execution efficiency, and memory-construction cost.

cs.CL↗

arXiv2026

LiveServe: Interaction-Aware Serving for Real-Time Omni-Modal LLMs

Realtime omni-modal LMs support speech-centric conversations where users stream inputs, hear generated audio, and interrupt freely. Existing Omni-LM serving systems still rely on throughput-oriented LLM scheduling and LRU KV offloading. These policies ignore audio playback and multi-turn reuse: they may generate tokens far beyond what users hear, wasting work after barge-in, and evict KV state needed in the next turn. LiveServe is an interaction-aware serving system for realtime Omni-LM interaction. It exposes playback progress, speech activity, and barge-in events to the serving pipeline. The scheduler prioritizes first-audio and near-underrun sessions while limiting generation beyond the playback frontier. The KV manager uses next-use-aware eviction and preloads likely-needed KV during user speech to hide reload latency. On vLLM-Omni, LiveServe improves realtime serving across two Omni-LMs and mixed workloads. It lowers P90 audio TTFP by $1.55\times$ on average and up to $2.21\times$, while improving completed-request throughput by $1.15\times$ on average and up to $1.56\times$, and moves most KV reload work off the next-turn critical path.

cs.DC↗

arXiv2026

Text-attributed Graph Condensation via Text Selection and Attribute Matching

Text-Attributed Graph (TAG) is an important type of graph structured data, where each node has a text description. TAG models usually train a Graph Neural Network (GNN) and language model jointly, which leads to high space and time consumption, especially on large datasets. To mitigate this, we propose TAGSAM, a condensation method that compresses TAGs while preserving training accuracy. TAGSAM comes with two key designs, i.e., subgraph text Selection and Attribute similarity Matching, which compress the text description and graph topology of TAG, respectively. For the texts, subgraph text selection selects and merges representative text chunks from multiple related text descriptions by maximizing mutual information. For the graph topology, popular condensation methods based on Matching Training Trajectories (MTT) suffer from high variance, which hinders accuracy. Our attribute similarity matching mitigates this issue by aligning stable similarity matrices. We evaluate TAGSAM against six state-of-the-art baselines, where it showcases superior performance. For the same compressed size, TAGSAM improves upon the best-performing baseline by an average of 4.9% in accuracy. Furthermore, it maintains competitive training accuracy even when the TAG is condensed to just 1% size. Our code is available at https://github.com/SundayVHan/TAGSAM

cs.LG↗

arXiv2026

MiCU: End-to-End Smart Home Command Understanding with Large Language Model

Command understanding systems in smart home ecosystems can automate device control and substantially improve user experience. However, while they perform well on precise utterances (e.g., "turn on the bedroom light"), they struggle with ambiguous or misaligned commands (e.g., "make the bedroom cozy"). Large language models (LLMs) generalize well across various domains and can outperform traditional rule-based systems on such tasks, but their effectiveness is often constrained by scarce domain-specific data, insufficient task-specific adaptation, and high computational costs. In this paper, we propose an automated training data synthesis workflow using user logs and LLMs; then we build MiCU, a domain-specific LLM that excels at command understanding. Specifically, we employ curriculum learning to inject domain knowledge into the base LLM, then we enhance its reasoning ability via cold-start training combined with reinforcement learning (RL) guided by domain-specific thinking rules. Additionally, we introduce a token compression technique that condenses device description into a single special token, substantially reducing inference overhead and enabling \model-fast, an efficient variant optimized for long inputs. Extensive experiments show that MiCU significantly outperforms baselines, with an average accuracy gain of 20.01% across all device categories. We have deployed MiCU in the Xiaomi Home app, receiving approximately 1.7 million page views per day. Production evaluations show that MiCU reduces user correction rate by 1.57% and increases human audited accuracy by 32.05%. Our data and code are available at https://github.com/xiaomi-research/iot_spec_llm

cs.CL↗

arXiv2026

RecMem: Recurrence-based Memory Consolidation for Efficient and Effective Long-Running LLM Agents

Memory systems often organize user-agent interactions as retrievable external memory and are crucial for long-running agents by overcoming the limited context windows of LLMs. However, existing memory systems invoke LLMs to process every incoming interaction for memory extraction, and such an eager memory consolidation scheme leads to substantial token consumption. To tackle this problem, we propose RecMem by rethinking when memory consolidation should be conducted. RecMem stores incoming interactions in a subconscious memory layer and encode them using lightweight embedding models for retrieval. LLMs are only invoked to extract episodic and semantic memory when sustained recurrence are observed for semantically similar interactions. Such recurrence-based consolidation works because these interactions correspond to a semantic cluster with rich information and thus are worth extraction and summarization. To improve accuracy, RecMem also incorporates a semantic refinement mechanism that recovers the fine-grained facts omitted by memory extraction. Experiments show that RecMem reduces the memory construction token cost of three SOTA memory systems by up to 87% while exceeding their accuracy.

cs.CL↗

arXiv2026

BatchWeave: A Consistent Object-Store-Native Data Plane for Large Foundation Model Training

Modern Large Foundation Model (LFM) training has transformed the data pipeline from a static ingestion layer into a dynamic component that must co-evolve with the training process. Existing systems are ill-equipped: colocated dataloaders offer no failure isolation, while message queue-based disaggregated dataloaders operate on a record/offset abstraction that cannot express the batch-level semantics required by distributed training. We present BatchWeave, an object-store-native training data plane for distributed LFM training. BatchWeave uses versioned manifests and conditional object writes to coordinate batch publication, recovery, and lifecycle management. First, it introduces the Transactional Global Batch (TGB), which builds on versioned-manifest ACID storage semantics and extends them with training-specific consistency, including atomic all-rank batch visibility, a globally ordered step sequence, checkpoint-aligned lifecycle management, and end-to-end exactly-once recovery. Second, it realizes recovery and retention directly in the storage layer, by durably persisting producer state through the commit protocol and tying reclamation to distributed checkpoint state. Third, its Decentralized Adaptive Commit (DAC) algorithm sustains stable ingestion throughput as the manifest grows, without any inter-producer communication. Evaluations on large-scale multimodal pre-training and SFT workloads using 64 GPUs show that BatchWeave outperforms colocated dataloader throughput while providing full failure isolation, outperforms Apache Kafka in ingestion throughput, and achieves lower consumer read latency than Kafka.

cs.DC↗

arXiv2026

UniVA: Unified Value Alignment for Generative Recommendation in Online Advertising at Tencent

Generative Recommendation (GR) reformulates recommendation as next-token generation over item Semantic IDs (SIDs) and has shown promise in industrial applications. However, extending GR to advertising is non-trivial, as advertising recommendation must jointly account for user relevance and commercial value. This creates a mismatch: high generation likelihood does not necessarily imply high advertising utility. As a result, valuable ads may be poorly distinguished in the SID space, pruned during autoregressive decoding, or missed when request-invalid branches consume limited beam capacity during online serving. To address this problem, we propose UniVA, a Unified Value Alignment framework for generative advertising recommendation. UniVA aligns commercial value across the entire pipeline of SID construction, autoregressive decoding, and online serving. Commercial SID Tokenization injects business attributes and bid information into SID construction. A Generation-as-Ranking SID Decoder then fuses generation scores with token-level value estimates during autoregressive decoding. {Finally, Value-Aware Constrained Serving restricts the fused decoding process to request-valid SID paths through a personalized trie.} Experiments on the Tencent WeChat Channels advertising platform show that UniVA achieves a 37.04\% relative improvement in offline Hit Rate@100 over the baseline and lifts gross merchandise value (GMV) by 1.5\% in online A/B tests.

cs.IR↗

arXiv2026

Efficient Serving for Dynamic Agent Workflows with Prediction-based KV-Cache Management

LLM-based workflows compose specialized agents to execute complex tasks, and these agents usually share substantial context, allowing KV-Cache reuse to save computation. Existing approaches either manage KV-Cache at agent level and fail to exploit the reuse opportunities within workflows, or manage cache at the workflow level but assume that each workflow calls a static sequence of agents. However, practical workflows are typically dynamic, where the sequence of invoked agents and thus induced cache reuse opportunities depend on the context of each task. To serve such dynamic workflows efficiently, we build a system dubbed PBKV (\textbf{P}rediction-\textbf{B}ased \textbf{KV}-Cache Management). For each workflow, PBKV predicts the agent invocations in several future steps by fusing the guidance from historical workflows and context of the target workflow. Based on the predictions, PBKV estimates the reuse potential of cache entries and keeps the high-potential entries in GPU memory. To be robust to prediction errors, PBKV utilizes the predictions conservatively during both cache eviction and prefetching. Experiments on three workflow benchmarks show that PBKV achieves up to $1.85\times$ speedup over LRU on dynamic workflows, and up to $1.26\times$ speedup over the SOTA baseline KVFlow on the static workflow.

cs.LG↗

arXiv2026

MTServe: Efficient Serving for Generative Recommendation Models with Hierarchical Caches

Generative recommendation (GR) offers superior modeling capabilities but suffers from prohibitive inference costs due to the repeated encoding of long user histories. While cross-request Key-Value (KV) cache reuse presents a significant optimization opportunity, the massive scale of individual user states creates a storage explosion that far exceeds physical GPU limits. We propose MTServe, a hierarchical cache management system that virtualizes GPU memory by leveraging host RAM as a scalable backup store. To bridge the I/O gap between tiers, MTServe introduces a suite of system-level optimizations, including a hybrid storage layout, an asynchronous data transfer pipeline, and a locality-driven replacement policy. On both public and production datasets, MTServe delivers up to 3.1* speedup while maintaining near-perfect hit ratios (>98.5%).

cs.LG↗

arXiv2026

Scheduling LLM Inference with Uncertainty-Aware Output Length Predictions

To schedule LLM inference, the \textit{shortest job first} (SJF) principle is favorable by prioritizing requests with short output lengths to avoid head-of-line (HOL) blocking. Existing methods usually predict a single output length for each request to facilitate scheduling. We argue that such a \textit{point estimate} does not match the \textit{stochastic} decoding process of LLM inference, where output length is \textit{uncertain} by nature and determined by when the end-of-sequence (EOS) token is sampled. Hence, the output length of each request should be fitted with a distribution rather than a single value. With an in-depth analysis of empirical data and the stochastic decoding process, we observe that output length follows a heavy-tailed distribution and can be fitted with the log-t distribution. On this basis, we propose a simple metric called Tail Inflated Expectation (TIE) to replace the output length in SJF scheduling, which adjusts the expectation of a log-t distribution with its tail probabilities to account for the risk that a request generates long outputs. To evaluate our TIE scheduler, we compare it with three strong baselines, and the results show that TIE reduces the per-token latency by $2.31\times$ for online inference and improves throughput by $1.42\times$ for offline data generation.

cs.LG↗

arXiv2026

FuxiShuffle: An Adaptive and Resilient Shuffle Service for Distributed Data Processing on Alibaba Cloud

Shuffle exchanges intermediate results between upstream and downstream operators in distributed data processing and is usually the bottleneck due to factors such as small random I/Os and network contention. Several systems have been designed to improve shuffle efficiency, but from our experiences of running ultra-large clusters at Alibaba Cloud MaxCompute platform, we observe that they can not adapt to highly dynamic job characteristics and cluster resource conditions, and their fault tolerance mechanisms are passive and inefficient when failures are inevitable. To tackle their limitations, we design and implement FuxiShuffle as a general data shuffle service for the ultra-large production environment of MaxCompute, featuring good adaptability and efficient failure resilience. Specifically, to achieve good adaptability, FuxiShuffle dynamically selects the shuffle mode based on runtime information, conducts progress-aware scheduling for the downstream workers, and automatically determines the most suitable backup strategy for each shuffle data chunk. To make failure resilience efficient, FuxiShuffle actively ensures data availability with multi-replica failover, prevents memory overflow with careful memory management, and employs an incremental recovery mechanism that does not lose computation progress. Our experiments show that, compared to baseline systems, FuxiShuffle significantly reduces not only end-to-end job completion time but also aggregate resource consumption. Micro experiments suggest that our designs are effective in improving adaptability and failure resilience.

cs.DC↗

arXiv2026

End-to-End Semantic ID Generation for Generative Advertisement Recommendation

Generative Recommendation (GR) has excelled by framing recommendation as next-token prediction. This paradigm relies on Semantic IDs (SIDs) to tokenize large-scale items into discrete sequences. Existing GR approaches predominantly generate SIDs via Residual Quantization (RQ), where items are encoded into embeddings and then quantized to discrete SIDs. However, this paradigm suffers from inherent limitations: 1) Objective misalignment and semantic degradation stemming from the two-stage compression; 2) Error accumulation inherent in the structure of RQ. To address these limitations, we propose UniSID, a Unified SID generation framework for generative advertisement recommendation. Specifically, we jointly optimize embeddings and SIDs in an end-to-end manner from raw advertising data, enabling semantic information to flow directly into the SID space and thus addressing the inherent limitations of the two-stage cascading compression paradigm. To capture fine-grained semantics, a multi-granularity contrastive learning strategy is introduced to align distinct items across SID levels. Finally, a summary-based ad reconstruction mechanism is proposed to encourage SIDs to capture high-level semantic information that is not explicitly present in advertising contexts. Experiments demonstrate that UniSID consistently outperforms state-of-the-art SID generation methods, yielding up to a 4.62% improvement in Hit Rate metrics across downstream advertising scenarios compared to the strongest baseline.

cs.IR↗

arXiv2025

CUBE: A Cardinality Estimator Based on Neural CDF

Modern database optimizer relies on cardinality estimator, whose accuracy directly affects the optimizer's ability to choose an optimal execution plan. Recent work on data-driven methods has leveraged probabilistic models to achieve higher estimation accuracy, but these approaches cannot guarantee low inference latency at the same time and neglect scalability. As data dimensionality grows, optimization time can even exceed actual query execution time. Furthermore, inference with probabilistic models by sampling or integration procedures unpredictable estimation result and violate stability, which brings unstable performance with query execution and make database tuning hard for database users. In this paper, we propose a novel approach to cardinality estimation based on cumulative distribution function(CDF), which calculates range query cardinality without sampling or integration, ensuring accurate and predictable estimation results. With inference acceleration by merging calculations, we can achieve fast and nearly constant inference speed while maintaining high accuracy, even as dimensionality increases, which is over 10x faster than current state-of-the-art data-driven cardinality estimator. This demonstrates its excellent dimensional scalability, making it well-suited for real-world database applications.

cs.DB↗

arXiv2025

DevPiolt: Operation Recommendation for IoT Devices at Xiaomi Home

Operation recommendation for IoT devices refers to generating personalized device operations for users based on their context, such as historical operations, environment information, and device status. This task is crucial for enhancing user satisfaction and corporate profits. Existing recommendation models struggle with complex operation logic, diverse user preferences, and sensitive to suboptimal suggestions, limiting their applicability to IoT device operations. To address these issues, we propose DevPiolt, a LLM-based recommendation model for IoT device operations. Specifically, we first equip the LLM with fundamental domain knowledge of IoT operations via continual pre-training and multi-task fine-tuning. Then, we employ direct preference optimization to align the fine-tuned LLM with specific user preferences. Finally, we design a confidence-based exposure control mechanism to avoid negative user experiences from low-quality recommendations. Extensive experiments show that DevPiolt significantly outperforms baselines on all datasets, with an average improvement of 69.5% across all metrics. DevPiolt has been practically deployed in Xiaomi Home app for one quarter, providing daily operation recommendations to 255,000 users. Online experiment results indicate a 21.6% increase in unique visitor device coverage and a 29.1% increase in page view acceptance rates.

cs.AI↗