arXiv subjects

Yao Liu

Publications and source records attributed to Yao Liu.

At least 73 records · Page 4Linked to original sources

arXiv2025

Network and Systems Performance Characterization of MCP-Enabled LLM Agents

Model Context Protocol (MCP) has recently gained increased attention within the AI community for providing a standardized way for large language models (LLMs) to interact with external tools and services, significantly enhancing their capabilities. However, the inclusion of extensive contextual information, including system prompts, MCP tool definitions, and context histories, in MCP-enabled LLM interactions, dramatically inflates token usage. Given that LLM providers charge based on tokens, these expanded contexts can quickly escalate monetary costs and increase the computational load on LLM services. This paper presents a comprehensive measurement-based analysis of MCP-enabled interactions with LLMs, revealing trade-offs between capability, performance, and cost. We explore how different LLM models and MCP configurations impact key performance metrics such as token efficiency, monetary cost, task completion times, and task success rates, and suggest potential optimizations, including enabling parallel tool calls and implementing robust task abort mechanisms. These findings provide useful insights for developing more efficient, robust, and cost-effective MCP-enabled workflows.

cs.DC↗

arXiv2025

WebAgent-R1: Training Web Agents via End-to-End Multi-Turn Reinforcement Learning

While reinforcement learning (RL) has demonstrated remarkable success in enhancing large language models (LLMs), it has primarily focused on single-turn tasks such as solving math problems. Training effective web agents for multi-turn interactions remains challenging due to the complexity of long-horizon decision-making across dynamic web interfaces. In this work, we present WebAgent-R1, a simple yet effective end-to-end multi-turn RL framework for training web agents. It learns directly from online interactions with web environments by asynchronously generating diverse trajectories, entirely guided by binary rewards depending on task success. Experiments on the WebArena-Lite benchmark demonstrate the effectiveness of WebAgent-R1, boosting the task success rate of Qwen-2.5-3B from 6.1% to 33.9% and Llama-3.1-8B from 8.5% to 44.8%, significantly outperforming existing state-of-the-art methods and strong proprietary models such as OpenAI o3. In-depth analyses reveal the effectiveness of the thinking-based prompting strategy and test-time scaling through increased interactions for web tasks. We further investigate different RL initialization policies by introducing two variants, namely WebAgent-R1-Zero and WebAgent-R1-CoT, which highlight the importance of the warm-up training stage (i.e., behavior cloning) and provide insights on incorporating long chain-of-thought (CoT) reasoning in web agents.

cs.CL↗

arXiv2025

Improving Sampling Efficiency in RLVR through Adaptive Rollout and Response Reuse

Large language models (LLMs) have achieved impressive reasoning performance, with reinforcement learning with verifiable rewards (RLVR) emerging as a standard paradigm for post-training. A representative algorithm, group relative policy optimization (GRPO) (Shao et al., 2024), computes advantages by normalizing outcome rewards within response groups, but suffers from a vanishing advantage issue when all responses in a group receive identical rewards. To address this issue, we propose Adaptive Rollout and Response Reuse Policy Optimization (AR3PO), a sampling efficient RLVR algorithm that introduces two novel techniques: adaptive rollout, which dynamically allocates more responses to difficult prompts while saving computation on easier ones, and response reuse, which leverages previously generated correct responses to provide useful training signals. We compare AR3PO with strong RLVR baselines on multiple representative benchmarks using two different families of base models. Across the 7B and 8B models, AR3PO consistently outperforms GRPO and matches or surpasses DAPO (Yu et al., 2025), reducing rollout cost by up to 4.2x. On the larger 32B model, AR3PO achieves comparable performance to DAPO at similar training steps while maintaining substantially lower rollout cost.

cs.LG↗

arXiv2025

When Your Reviewer is an LLM: Biases, Divergence, and Prompt Injection Risks in Peer Review

Peer review is the cornerstone of academic publishing, yet the process is increasingly strained by rising submission volumes, reviewer overload, and expertise mismatches. Large language models (LLMs) are now being used as "reviewer aids," raising concerns about their fairness, consistency, and robustness against indirect prompt injection attacks. This paper presents a systematic evaluation of LLMs as academic reviewers. Using a curated dataset of 1,441 papers from ICLR 2023 and NeurIPS 2022, we evaluate GPT-5-mini against human reviewers across ratings, strengths, and weaknesses. The evaluation employs structured prompting with reference paper calibration, topic modeling, and similarity analysis to compare review content. We further embed covert instructions into PDF submissions to assess LLMs' susceptibility to prompt injection. Our findings show that LLMs consistently inflate ratings for weaker papers while aligning more closely with human judgments on stronger contributions. Moreover, while overarching malicious prompts induce only minor shifts in topical focus, explicitly field-specific instructions successfully manipulate specific aspects of LLM-generated reviews. This study underscores both the promises and perils of integrating LLMs into peer review and points to the importance of designing safeguards that ensure integrity and trust in future review processes.

cs.CY↗

arXiv2025

Can Large Language Models Act as Ensembler for Multi-GNNs?

Graph Neural Networks (GNNs) have emerged as powerful models for learning from graph-structured data. However, GNNs lack the inherent semantic understanding capability of rich textual node attributes, limiting their effectiveness in applications. On the other hand, we empirically observe that for existing GNN models, no one can consistently outperforms others across diverse datasets. In this paper, we study whether LLMs can act as an ensembler for multi-GNNs and propose the LensGNN model. The model first aligns multiple GNNs, mapping the representations of different GNNs into the same space. Then, through LoRA fine-tuning, it aligns the space between the GNN and the LLM, injecting graph tokens and textual information into LLMs. This allows LensGNN to ensemble multiple GNNs and take advantage of the strengths of LLM, leading to a deeper understanding of both textual semantic information and graph structural information. The experimental results show that LensGNN outperforms existing models. This research advances text-attributed graph ensemble learning by providing a robust and superior solution for integrating semantic and structural information. We provide our code and data here: https://github.com/AquariusAQ/LensGNN.

cs.AI↗

arXiv2025

Ergodicity of conditional McKean-Vlasov jump diffusions

In this paper, we are interested in conditional McKean-Vlasov jump diffusions, which are also termed as McKean-Vlasov stochastic differential equations with jump idiosyncratic noise and jump common noise. As far as conditional McKean-Vlasov jump diffusions are concerned, the corresponding conditional distribution flow is a measure-valued process, which indeed satisfies a stochastic partial integral differential equation driven by a Poisson random measure. Via a novel construction of the asymptotic coupling by reflection, we explore the ergodicity of the underlying measure-valued process corresponding to a one-dimensional conditional McKean-Vlasov jump diffusion when the associated drift term fulfils a partially dissipative condition with respect to the spatial variable. In addition, the theory derived demonstrates that the intensity of the jump common noise and the jump idiosyncratic noise can simultaneously enhance the convergence rate of the exponential ergodicity.

math.PR↗

arXiv2025

MMAU-Pro: A Challenging and Comprehensive Benchmark for Holistic Evaluation of Audio General Intelligence

Audio comprehension-including speech, non-speech sounds, and music-is essential for achieving human-level intelligence. Consequently, AI agents must demonstrate holistic audio understanding to qualify as generally intelligent. However, evaluating auditory intelligence comprehensively remains challenging. To address this gap, we introduce MMAU-Pro, the most comprehensive and rigorously curated benchmark for assessing audio intelligence in AI systems. MMAU-Pro contains 5,305 instances, where each instance has one or more audios paired with human expert-generated question-answer pairs, spanning speech, sound, music, and their combinations. Unlike existing benchmarks, MMAU-Pro evaluates auditory intelligence across 49 unique skills and multiple complex dimensions, including long-form audio comprehension, spatial audio reasoning, multi-audio understanding, among others. All questions are meticulously designed to require deliberate multi-hop reasoning, including both multiple-choice and open-ended response formats. Importantly, audio data is sourced directly ``from the wild" rather than from existing datasets with known distributions. We evaluate 22 leading open-source and proprietary multimodal AI models, revealing significant limitations: even state-of-the-art models such as Gemini 2.5 Flash and Audio Flamingo 3 achieve only 59.2% and 51.7% accuracy, respectively, approaching random performance in multiple categories. Our extensive analysis highlights specific shortcomings and provides novel insights, offering actionable perspectives for the community to enhance future AI systems' progression toward audio general intelligence. The benchmark and code is available at https://sonalkum.github.io/mmau-pro.

eess.AS↗

arXiv2025

MINDS. Strong oxygen depletion in the inner regions of a very low-mass star disk?

JWST is discovering a plethora of species in planet-forming disks around very low-mass stars such as C2H2, C6H6, C4H2, CH3 etc. The column densities of these species retrieved from 0D slab models are very large, e.g. of the order of $10^{20}$\,cm$^{-2}$. This is indicating a carbon-dominated chemistry in a gas with a high C/O ratio. The disk around 2MASS-J1605321-1993159 (M4.5) is one such source showing a molecular pseudo-continuum of C2H2. Still two oxygen-bearing molecules, CO and CO2 are also detected in this source. We aim to take the next step beyond 0D slab models to interpret the spectrum. We examine whether 2D thermo-chemical disk models can produce the large inferred column densities of \ce{C2H2} in the inner regions of the disk and produce a pseudo-continuum in the mid-IR spectrum. We also want to constrain whether depletion of oxygen or enrichment of carbon is causing the high C/O ratio triggering a carbon-dominated chemistry. We utilize the radiative thermo-chemical disk model P{\tiny RO}D{\tiny I}M{\tiny O} to identify a disk structure which is capable of producing the observed molecular emission of species such as CO, CO2, C2H2, and H2O simultaneously. The spectrum is generated using the fast line tracer FLiTs. We derive the gas temperature $\langle T \rangle$, column density $\langle$ log$_{\rm {10}} N\rangle$ and the emitting area $\langle r_{\rm{1}} - r_{\rm{2}} \rangle$ for these molecules from the 2D disk model and compare them to the parameters retrieved originally from 0D slab models. We use the different effect that changing the O or C abundance has on CO and C2H2 respectively to discriminate between O depletion and C enhancement. We find that a disk structure characterised by the presence of a gap can best explain the observations. The inner disk is strongly depleted in dust, especially small grains ($<5\,μ$m), and ..

astro-ph.EP↗

arXiv2025

DRKF: Decoupled Representations with Knowledge Fusion for Multimodal Emotion Recognition

Multimodal emotion recognition (MER) aims to identify emotional states by integrating and analyzing information from multiple modalities. However, inherent modality heterogeneity and inconsistencies in emotional cues remain key challenges that hinder performance. To address these issues, we propose a Decoupled Representations with Knowledge Fusion (DRKF) method for MER. DRKF consists of two main modules: an Optimized Representation Learning (ORL) Module and a Knowledge Fusion (KF) Module. ORL employs a contrastive mutual information estimation method with progressive modality augmentation to decouple task-relevant shared representations and modality-specific features while mitigating modality heterogeneity. KF includes a lightweight self-attention-based Fusion Encoder (FE) that identifies the dominant modality and integrates emotional information from other modalities to enhance the fused representation. To handle potential errors from incorrect dominant modality selection under emotionally inconsistent conditions, we introduce an Emotion Discrimination Submodule (ED), which enforces the fused representation to retain discriminative cues of emotional inconsistency. This ensures that even if the FE selects an inappropriate dominant modality, the Emotion Classification Submodule (EC) can still make accurate predictions by leveraging preserved inconsistency information. Experiments show that DRKF achieves state-of-the-art (SOTA) performance on IEMOCAP, MELD, and M3ED. The source code is publicly available at https://github.com/PANPANKK/DRKF.

cs.MM↗

arXiv2025

Improving Graph Out-of-distribution Generalization Beyond Causality

Existing methods for graph out-of-distribution (OOD) generalization primarily rely on empirical studies on synthetic datasets. Such approaches tend to overemphasize the causal relationships between invariant sub-graphs and labels, thereby neglecting the non-negligible role of environment in real-world scenarios. In contrast to previous studies that impose rigid independence assumptions on environments and invariant sub-graphs, this paper presents the theorems of environment-label dependency and mutable rationale invariance, where the former characterizes the usefulness of environments in determining graph labels while the latter refers to the mutable importance of graph rationales. Based on analytic investigations, a novel variational inference based method named ``Probability Dependency on Environments and Rationales for OOD Graphs on Real-world Data'' (DEROG) is introduced. To alleviate the adverse effect of unknown prior knowledge on environments and rationales, DEROG utilizes generalized Bayesian inference. Further, DEROG employs an EM-based algorithm for optimization. Finally, extensive experiments on real-world datasets under different distribution shifts are conducted to show the superiority of DEROG. Our code is publicly available at https://github.com/LEOXC1571/DEROG.

cs.LG↗

arXiv2025

Weakly-supervised VLM-guided Partial Contrastive Learning for Visual Language Navigation

Visual Language Navigation (VLN) is a fundamental task within the field of Embodied AI, focusing on the ability of agents to navigate complex environments based on natural language instructions. Despite the progress made by existing methods, these methods often present some common challenges. First, they rely on pre-trained backbone models for visual perception, which struggle with the dynamic viewpoints in VLN scenarios. Second, the performance is limited when using pre-trained LLMs or VLMs without fine-tuning, due to the absence of VLN domain knowledge. Third, while fine-tuning LLMs and VLMs can improve results, their computational costs are higher than those without fine-tuning. To address these limitations, we propose Weakly-supervised Partial Contrastive Learning (WPCL), a method that enhances an agent's ability to identify objects from dynamic viewpoints in VLN scenarios by effectively integrating pre-trained VLM knowledge into the perception process, without requiring VLM fine-tuning. Our method enhances the agent's ability to interpret and respond to environmental cues while ensuring computational efficiency. Experimental results have shown that our method outperforms the baseline methods on multiple benchmarks, which validate the effectiveness, robustness and generalizability of our method.

cs.CV↗

arXiv2025

Fundamental Limits of Multiple-Access Integrated Sensing and Communication Systems

A state-dependent discrete memoryless multiple access channel is considered to model an integrated sensing and communication system, where two transmitters wish to convey messages to a receiver while simultaneously estimating the state parameter sequences through echo signals. In particular, the sensing state parameters are assumed to be correlated with the channel state. In this setup, improved inner and outer bounds for capacity-distortion region are derived. The inner bound is based on an achievable scheme that combines message cooperation and joint compression of past transmitted codewords and echo signals at each transmitter, resulting in unified cooperative communication and sensing. The outer bound is based on the ideas of dependence balance for communication rate, genie-aided state estimator and rate-limited constraints on sensing distortion. The proposed inner and outer bounds are proved to improve the state-of-the-art bounds. Finally, numerical examples are provided to demonstrate that our new inner and outer bounds strictly improve the existing results.

cs.IT↗

arXiv2025

EyeNavGS: A 6-DoF Navigation Dataset and Record-n-Replay Software for Real-World 3DGS Scenes in VR

3D Gaussian Splatting (3DGS) is an emerging media representation that reconstructs real-world 3D scenes in high fidelity, enabling 6-degrees-of-freedom (6-DoF) navigation in virtual reality (VR). However, developing and evaluating 3DGS-enabled applications and optimizing their rendering performance, require realistic user navigation data. Such data is currently unavailable for photorealistic 3DGS reconstructions of real-world scenes. This paper introduces EyeNavGS (EyeNavGS), the first publicly available 6-DoF navigation dataset featuring traces from 46 participants exploring twelve diverse, real-world 3DGS scenes. The dataset was collected at two sites, using the Meta Quest Pro headsets, recording the head pose and eye gaze data for each rendered frame during free world standing 6-DoF navigation. For each of the twelve scenes, we performed careful scene initialization to correct for scene tilt and scale, ensuring a perceptually-comfortable VR experience. We also release our open-source SIBR viewer software fork with record-and-replay functionalities and a suite of utility tools for data processing, conversion, and visualization. The EyeNavGS dataset and its accompanying software tools provide valuable resources for advancing research in 6-DoF viewport prediction, adaptive streaming, 3D saliency, and foveated rendering for 3DGS scenes. The EyeNavGS dataset is available at: https://symmru.github.io/EyeNavGS/.

cs.MM↗

arXiv2025

Enhancing LLM-based Hatred and Toxicity Detection with Meta-Toxic Knowledge Graph

The rapid growth of social media platforms has raised significant concerns regarding online content toxicity. When Large Language Models (LLMs) are used for toxicity detection, two key challenges emerge: 1) the absence of domain-specific toxic knowledge leads to false negatives; 2) the excessive sensitivity of LLMs to toxic speech results in false positives, limiting freedom of speech. To address these issues, we propose a novel method called MetaTox, leveraging graph search on a meta-toxic knowledge graph to enhance hatred and toxicity detection. First, we construct a comprehensive meta-toxic knowledge graph by utilizing LLMs to extract toxic information through a three-step pipeline, with toxic benchmark datasets serving as corpora. Second, we query the graph via retrieval and ranking processes to supplement accurate, relevant toxic knowledge. Extensive experiments and in-depth case studies across multiple datasets demonstrate that our MetaTox significantly decreases the false positive rate while boosting overall toxicity detection performance. Our code is available at https://github.com/YiboZhao624/MetaTox.

cs.CL↗

arXiv2025

AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents

Autonomy via agents using large language models (LLMs) for personalized, standardized tasks boosts human efficiency. Automating web tasks (like booking hotels within a budget) is increasingly sought after. Fulfilling practical needs, the web agent also serves as an important proof-of-concept example for various agent grounding scenarios, with its success promising advancements in many future applications. Prior research often handcrafts web agent strategies (e.g., prompting templates, multi-agent systems, search methods, etc.) and the corresponding in-context examples, which may not generalize well across all real-world scenarios. On the other hand, there has been limited study on the misalignment between a web agent's observation/action representation and the pre-training data of the LLM it's based on. This discrepancy is especially notable when LLMs are primarily trained for language completion rather than tasks involving embodied navigation actions and symbolic web elements. Our study enhances an LLM-based web agent by simply refining its observation and action space to better align with the LLM's capabilities. This approach enables our base agent to significantly outperform previous methods on a wide variety of web tasks. Specifically, on WebArena, a benchmark featuring general-purpose web interaction tasks, our agent AgentOccam surpasses the previous state-of-the-art and concurrent work by 9.8 (+29.4%) and 5.9 (+15.8%) absolute points respectively, and boosts the success rate by 26.6 points (+161%) over similar plain web agents with its observation and action space alignment. We achieve this without using in-context examples, new agent roles, online feedback or search strategies. AgentOccam's simple design highlights LLMs' impressive zero-shot performance on web tasks, and underlines the critical role of carefully tuning observation and action spaces for LLM-based agents.

cs.AI↗

arXiv2025

Invisible Prompts, Visible Threats: Malicious Font Injection in External Resources for Large Language Models

Large Language Models (LLMs) are increasingly equipped with capabilities of real-time web search and integrated with protocols like Model Context Protocol (MCP). This extension could introduce new security vulnerabilities. We present a systematic investigation of LLM vulnerabilities to hidden adversarial prompts through malicious font injection in external resources like webpages, where attackers manipulate code-to-glyph mapping to inject deceptive content which are invisible to users. We evaluate two critical attack scenarios: (1) "malicious content relay" and (2) "sensitive data leakage" through MCP-enabled tools. Our experiments reveal that indirect prompts with injected malicious font can bypass LLM safety mechanisms through external resources, achieving varying success rates based on data sensitivity and prompt design. Our research underscores the urgent need for enhanced security measures in LLM deployments when processing external content.

cs.CR↗

arXiv2025

D3: Diversity, Difficulty, and Dependability-Aware Data Selection for Sample-Efficient LLM Instruction Tuning

Recent advancements in instruction tuning for large language models (LLMs) suggest that a small, high-quality dataset can significantly equip LLMs with instruction-following capabilities, outperforming large datasets often burdened by quality and redundancy issues. However, the challenge lies in automatically identifying valuable subsets from large datasets to boost both the effectiveness and efficiency of instruction tuning. In this paper, we first establish data selection criteria based on three distinct aspects of data value: diversity, difficulty, and dependability, and then propose the D3 method comprising two key steps of scoring and selection. Specifically, in the scoring step, we define the diversity function to measure sample distinctiveness and introduce the uncertainty-based prediction difficulty to evaluate sample difficulty by mitigating the interference of context-oriented generation diversity. Additionally, we integrate an external LLM for dependability assessment. In the selection step, we formulate the D3 weighted coreset objective, which jointly optimizes three aspects of data value to solve for the most valuable subset. The two steps of D3 can iterate multiple rounds, incorporating feedback to refine the selection focus adaptively. Experiments on both public datasets and the real-world Taobao Live application demonstrate the effectiveness of D3 in endowing LLMs with competitive or even superior instruction-following capabilities using less than 10\% of the entire dataset.

cs.LG↗

arXiv2025

From Demonstrations to Rewards: Alignment Without Explicit Human Preferences

One of the challenges of aligning large models with human preferences lies in both the data requirements and the technical complexities of current approaches. Predominant methods, such as RLHF, involve multiple steps, each demanding distinct types of data, including demonstration data and preference data. In RLHF, human preferences are typically modeled through a reward model, which serves as a proxy to guide policy learning during the reinforcement learning stage, ultimately producing a policy aligned with human preferences. However, in this paper, we propose a fresh perspective on learning alignment based on inverse reinforcement learning principles, where the optimal policy is still derived from reward maximization. However, instead of relying on preference data, we directly learn the reward model from demonstration data. This new formulation offers the flexibility to be applied even when only demonstration data is available, a capability that current RLHF methods lack, and it also shows that demonstration data offers more utility than what conventional wisdom suggests. Our extensive evaluation, based on public reward benchmark, HuggingFace Open LLM Leaderboard and MT-Bench, demonstrates that our approach compares favorably to state-of-the-art methods that rely solely on demonstration data.

cs.LG↗