arXiv subjects

Zhaoyu Hu

Publications and source records attributed to Zhaoyu Hu.

11 recordsLinked to original sources

arXiv2026

Tool Retrievers Are Underestimated: Annotation Expansion Reveals True Capability

In open-world scenarios with massive and evolving tool repositories, tool-augmented large language models rely on a retriever to surface relevant tools for a given query. Because such repositories often contain many tools that implement the same functionality, a single query can often be resolved by several distinct but functionally equivalent tool combinations, making the natural query-to-tool mapping inherently one-to-many. However, existing tool retrieval benchmarks annotate each query with a single relevant tool combination, collapsing this one-to-many mapping into a rigid one-to-one annotation and causing valid retrieved tools to be misjudged as failures. To address this, we propose ToolEX (Tool Equivalent eXpansion), a framework that automatically discovers and annotates the tool combinations functionally equivalent to the labeled ones. Applied to the 7,360-query Tool-DE benchmark, ToolEX finds that 67.9% of sub-queries admit equivalent alternatives, expanding the singular ground truth to an average of 5.3 valid combinations per query. Using the expanded benchmark ToolEQ, we re-evaluate eight base retrievers and two fine-tuned variants; metrics on ToolEQ rise substantially over Tool-DE, showing that one-to-one annotation systematically underestimates retrievers and that 30--47% of the reported fine-tuning gain is an evaluation artifact rather than genuine improvement. Applying the same pipeline to skill retrieval on SkillRet further confirms that the one-to-one problem extends beyond tool retrieval.

cs.SE↗

arXiv2026

ATLAS: Dual-Horizon Diagnostic Evaluation for Industrial Tool-Use Agents

Large language model (LLM) agents are increasingly deployed in user-facing services that require iterative tool use under dynamic business conditions. Reliable evaluation is essential for sustained improvement: it must reveal capability deficiencies, inform priorities, and assess interventions. Yet industrial agent service unfolds both through the iterative trajectory of a current request and through continued user interaction. Final-outcome assessment can therefore obscure where deficiencies arise and whether later service remains aligned with context from earlier exchanges. We propose ATLAS, a dual-horizon diagnostic evaluation framework for industrial tool-use agents. At the request horizon, trajectory-wise diagnostic signals relate deficiencies to execution locations and capability concerns. At the interaction horizon, user-wise signals assess whether service remains responsive across continued interaction. Together, these views provide structured diagnostic evidence for analyzing execution deficiencies and sustained service behavior. ATLAS instantiates them as executable signals with explicit evidence scopes and decision boundaries. LLM judge interfaces are calibrated against high-confidence references from real business logs; when needed, their decision behavior is distilled into efficient diagnostic models for lower-latency, lower-cost evaluation. The resulting feedback supports policy optimization. We evaluate ATLAS on Meituan Xiaotuan production traffic. Offline experiments assess diagnostic-signal fidelity and replay-based policy improvement, while online A/B experiments show concurrent gains in user engagement, downstream business outcomes, and sampled human-audit quality.

cs.AI↗

arXiv2026

UniRefiner: Teaching Pre-trained ViTs to Self-Dispose Dross via Contrastive Register

Representation learning with Vision Transformers (ViTs) has advanced rapidly, yet the utility of large-scale models in spatially sensitive tasks is hindered by spurious tokens. Prior efforts to mitigate this have been limited, often defining these artifacts narrowly, for example, as simple high-norm outliers. We argue that this scope is insufficient. For dense prediction tasks, we posit that any token failing to encode location-aligned semantics should be treated as a spurious artifact. This broader definition reveals a more complex problem, leading us to systematically categorize and characterize three fundamental types of spurious tokens that corrupt spatial representations. Based on this comprehensive diagnosis, we propose UniRefiner, a universal refinement framework that teaches pre-trained ViTs to self-dispose of these artifacts. UniRefiner uses contrastive registers to explicitly isolate and redistribute spurious tokens via a dual objective: (i) it aligns image tokens with filtered regular tokens to preserve semantics, and (ii) it aligns register tokens with detected spurious tokens to capture the spurious signals. Our method requires only a few epochs of fine-tuning on ~5k images to refine diverse ViTs, including massive models like EVA-CLIP-8B and InternViT-6B. Experiments demonstrate consistent and significant improvements: notably, the refined EVA-CLIP-8B achieves 51.9\% mIoU on ADE20K (+9.4\%), surpassing specialized vision models like DINOv2 (49.1\%), while zero-shot segmentation accuracy improves by up to 22\%. UniRefiner unlocks the latent spatial potential of existing large-scale foundation models, paving the way for their broader application.

cs.CV↗

arXiv2026

RecRM-Bench: Benchmarking Multidimensional Reward Modeling for Agentic Recommender Systems

The integration of Large Language Model (LLM) agents is transforming recommender systems from simple query-item matching towards deeply personalized and interactive recommendations. Reinforcement Learning (RL) provides an essential framework for the optimization of these agents in recommendation tasks. However, current methodologies remain limited by a reliance on single dimensional outcome-based rewards that focus exclusively on final user interactions, overlooking critical intermediate capabilities, such as instruction following and complex intent understanding. Despite the necessity for designing multi-dimensional reward, the field lacks a standardized benchmark to facilitate this development. To bridge this gap, we introduce RecRM-Bench, the largest and most comprehensive benchmark to date for agentic recommender systems. It comprises over 1 million structured entries across four core evaluation dimensions: instruction following, factual consistency, query-item relevance, and fine-grained user behavior prediction. By supporting comprehensive assessment from syntactic compliance to complex intent grounding and preference modeling, RecRM-Bench provides a foundational dataset for training sophisticated reward models. Furthermore, we propose a systematic framework for the construction of multi-dimensional reward models and the integration of a hybrid reward function, establishing a robust foundation for developing reliable and highly capable agentic recommender systems. The complete RecRM-Bench dataset is publicly available at https://huggingface.co/datasets/wwzeng/RecRM-Bench.

cs.IR↗

arXiv2026

XekRung Technical Report

We present XekRung, a frontier large language model for cybersecurity, designed to provide comprehensive security capabilities. To achieve this, we develop diverse data synthesis pipelines tailored to the cybersecurity domain, enabling the scalable construction of high-quality training data and providing a strong foundation for cybersecurity knowledge and understanding. Building on this foundation, we establish a complete training pipeline spanning continued pre-training (CPT), supervised fine-tuning (SFT), and reinforcement learning (RL) to further extend the model's capabilities. We further introduce a multi-dimensional evaluation system to guide the iterative improvement of both domain-specific and general-purpose abilities. Extensive experiments demonstrate that XekRung achieves state-of-the-art performance on cybersecurity-specific benchmarks among models of the same scale, while maintaining strong performance on general benchmarks.

cs.CR↗

arXiv2026

When Relevance Meets Novelty: Dual-Stable Periodic Optimization for Serendipitous Recommendation

Traditional recommendation systems tend to trap users in strong feedback loops by excessively pushing content aligned with their historical preferences, thereby limiting exploration opportunities and causing content fatigue. Although large language models (LLMs) demonstrate potential with their diverse content generation capabilities, existing LLM-enhanced dual-model frameworks face two major limitations: first, they overlook long-term preferences driven by group identity, leading to biased interest modeling; second, they suffer from static optimization flaws, as a one-time alignment process fails to leverage incremental user data for closed-loop optimization. To address these challenges, we propose the Co-Evolutionary Alignment (CoEA) method. For interest modeling bias, we introduce Dual-Stable Interest Exploration (DSIE) module, jointly modeling long-term group identity and short-term individual interests through parallel processing of behavioral sequences. For static optimization limitations, we design a Periodic Collaborative Optimization (PCO) mechanism. This mechanism regularly conducts preference verification on incremental data using the Relevance LLM, then guides the Novelty LLM to perform fine-tuning based on the verification results, and subsequently feeds back the output of the continually fine-tuned Novelty LLM to the Relevance LLM for re-evaluation, thereby achieving a dynamic closed-loop optimization. Extensive online and offline experiments verify the effectiveness of the CoEA model in serendipitous recommendation.

cs.IR↗

arXiv2025

Dynamic Forgetting and Spatio-Temporal Periodic Interest Modeling for Local-Life Service Recommendation

In the context of the booming digital economy, recommendation systems, as a key link connecting users and numerous services, face challenges in modeling user behavior sequences on local-life service platforms, including the sparsity of long sequences and strong spatio-temporal dependence. Such challenges can be addressed by drawing an analogy to the forgetting process in human memory. This is because users' responses to recommended content follow the recency effect and the cyclicality of memory. By exploring this, this paper introduces the forgetting curve and proposes Spatio-Temporal periodic Interest Modeling (STIM) with long sequences for local-life service recommendation. STIM integrates three key components: a dynamic masking module based on the forgetting curve, which is used to extract both recent spatiotemporal features and periodic spatiotemporal features; a query-based mixture of experts (MoE) approach that can adaptively activate expert networks under different dynamic masks, enabling the collaborative modeling of time, location, and items; and a hierarchical multi-interest network unit, which captures multi-interest representations by modeling the hierarchical interactions between the shallow and deep semantics of users' recent behaviors. By introducing the STIM method, we conducted online A/B tests and achieved a 1.54\% improvement in gross transaction volume (GTV). In addition, extended offline experiments also showed improvements. STIM has been deployed in a large-scale local-life service recommendation system, serving hundreds of millions of daily active users in core application scenarios.

cs.IR↗

arXiv2025

MTmixAtt: Integrating Mixture-of-Experts with Multi-Mix Attention for Large-Scale Recommendation

Industrial recommender systems critically depend on high-quality ranking models. However, traditional pipelines still rely on manual feature engineering and scenario-specific architectures, which hinder cross-scenario transfer and large-scale deployment. To address these challenges, we propose \textbf{MTmixAtt}, a unified Mixture-of-Experts (MoE) architecture with Multi-Mix Attention, designed for large-scale recommendation tasks. MTmixAtt integrates two key components. The \textbf{AutoToken} module automatically clusters heterogeneous features into semantically coherent tokens, removing the need for human-defined feature groups. The \textbf{MTmixAttBlock} module enables efficient token interaction via a learnable mixing matrix, shared dense experts, and scenario-aware sparse experts, capturing both global patterns and scenario-specific behaviors within a single framework. Extensive experiments on the industrial TRec dataset from Meituan demonstrate that MTmixAtt consistently outperforms state-of-the-art baselines including Transformer-based models, WuKong, HiFormer, MLP-Mixer, and RankMixer. At comparable parameter scales, MTmixAtt achieves superior CTR and CTCVR metrics; scaling to MTmixAtt-1B yields further monotonic gains. Large-scale online A/B tests validate the real-world impact: in the \textit{Homepage} scenario, MTmixAtt increases Payment PV by \textbf{+3.62\%} and Actual Payment GTV by \textbf{+2.54\%}. Overall, MTmixAtt provides a unified and scalable solution for modeling arbitrary heterogeneous features across scenarios, significantly improving both user experience and commercial outcomes.

cs.IR↗

arXiv2024

TASL-Net: Tri-Attention Selective Learning Network for Intelligent Diagnosis of Bimodal Ultrasound Video

In the intelligent diagnosis of bimodal (gray-scale and contrast-enhanced) ultrasound videos, medical domain knowledge such as the way sonographers browse videos, the particular areas they emphasize, and the features they pay special attention to, plays a decisive role in facilitating precise diagnosis. Embedding medical knowledge into the deep learning network can not only enhance performance but also boost clinical confidence and reliability of the network. However, it is an intractable challenge to automatically focus on these person- and disease-specific features in videos and to enable networks to encode bimodal information comprehensively and efficiently. This paper proposes a novel Tri-Attention Selective Learning Network (TASL-Net) to tackle this challenge and automatically embed three types of diagnostic attention of sonographers into a mutual transformer framework for intelligent diagnosis of bimodal ultrasound videos. Firstly, a time-intensity-curve-based video selector is designed to mimic the temporal attention of sonographers, thus removing a large amount of redundant information while improving computational efficiency of TASL-Net. Then, to introduce the spatial attention of the sonographers for contrast-enhanced video analysis, we propose the earliest-enhanced position detector based on structural similarity variation, on which the TASL-Net is made to focus on the differences of perfusion variation inside and outside the lesion. Finally, by proposing a mutual encoding strategy that combines convolution and transformer, TASL-Net possesses bimodal attention to structure features on gray-scale videos and to perfusion variations on contrast-enhanced videos. These modules work collaboratively and contribute to superior performance. We conduct a detailed experimental validation of TASL-Net's performance on three datasets, including lung, breast, and liver.

cs.CV↗

arXiv2020

GenCos' Behaviors Modeling Based on Q Learning Improved by Dichotomy

Q learning is widely used to simulate the behaviors of generation companies (GenCos) in an electricity market. However, existing Q learning method usually requires numerous iterations to converge, which is time-consuming and inefficient in practice. To enhance the calculation efficiency, a novel Q learning algorithm improved by dichotomy is proposed in this paper. This method modifies the update process of the Q table by dichotomizing the state space and the action space step by step. Simulation results in a repeated Cournot game show the effectiveness of the proposed algorithm.

eess.SY↗

arXiv2020

A Reinforcement Learning Method For Power Suppliers' Strategic Bidding with Insufficient Information

Power suppliers can exercise market power to gain higher profit. However, this becomes difficult when external information is extremely rare. To get a promising performance in an extremely incomplete information market environment, a novel model-free reinforcement learning algorithm based on the Learning Automata (LA) is proposed in this paper. Besides, this paper analyses the rationality and convergence of the algorithm in case studies based on the Cournot market model.

eess.SY↗