arXiv subjects

Yukai Yang

Publications and source records attributed to Yukai Yang.

9 recordsLinked to original sources

arXiv2026

Measuring Research Difficulty of Academic Papers: A Case Study in Natural Language Processing

With the rapid growth of the number of academic papers, systematically evaluating the difficulty of research and its relationship to academic impact offers important significance for research topic selection and resource allocation. However, current studies lack quantitative assessments of research difficulty and its correlation with academic impact. This paper proposes a comprehensive evaluation system for research difficulty, incorporating factors such as academic collaboration, content, and references. Taking the field of Natural Language Processing (NLP) as a case study, we extract both internal and external features from academic papers, compute multiple research difficulty indicators. We assign their weights using the entropy weight method and perform a weighted sum to obtain the research difficulty score of academic papers. This paper uses the citation frequency of academic papers to measure academic impact. To validate our approach, NLP experts assessed the difficulty of a sample of papers, and correlation analyses confirmed the reliability of our measurement. Empirical results reveal that in NLP, factors such as the number of pages, reference count, and participation of high-level institutions are significantly associated with academic impact. Moreover, we identify an inverted U-shaped relationship between research difficulty and academic impact. It suggests that moderately difficult research tends to achieve greater academic impact.

cs.DL↗

arXiv2026

Context-Fractured Decomposition Attacks on Tool-Using LLM Agents: Exploiting Artifact Provenance Gaps

Tool-using LLM agents interact with the world through actions that persist state in artifacts (e.g., workspace files or logs). Consequently, jailbreak defenses must reason about cross-step composition rather than isolated text. Yet most existing attacks and defenses, including ``multi-turn'' jailbreaks such as Crescendo and Tree of Attacks,still assume a single contiguous conversation visible to the defender. This assumption breaks down in real agent pipelines, where enforcement is fragmented across tools, modules, and time, and where artifact provenance is often not tracked. We operationalize a deployment failure mode for tool-using LLM agents, the \emph{provenance gap}, and study reproducible triggers for it: \emph{Context-Fractured Decomposition} (CFD), a family of cross-context multi-step jailbreaks that preserve benign-looking intermediate artifacts from an early interaction and elicit harmful behavior much later, potentially in a different agent instance or workflow stage, via individually innocuous tool actions whose risk emerges only under delayed artifact-mediated composition. We instrument the failure mode with trace-level diagnostics and outline a verifiable mitigation direction (provenance lineage tagging). Across agent-system jailbreak benchmarks, CFD improves success rates by up to 28.3 percentage points over state-of-the-art baselines, even against strong single-turn judges. Disclaimer: This paper contains examples of harmful or offensive language.

cs.CR↗

arXiv2026

SpikingMoE: SDPrompt-Guided Dynamic Expert Fusion in Spiking Neural Networks

Spiking Neural Networks (SNNs) provide an energy-efficient paradigm for visual recognition. We present SpikingMoE, which integrates a spike-driven Transformer with a Mixture-of-Experts (MoE) framework for dynamic computation. Inspired by the lateral geniculate nucleus (LGN), a spike-driven prompt (SDprompt) enables input-dependent expert routing in a biologically plausible manner. By replacing standard MLPs with spike-compatible expert modules and enforcing binary spike communication, SpikingMoE is designed for neuromorphic hardware. Experiments on CIFAR-10 and CIFAR-100 achieve 94.09% and 74.54% top-1 accuracy, showing that modular expert routing can be incorporated while retaining reasonable performance. To our knowledge, SpikingMoE is the first open-source SNN framework that integrates MoE into a spike-driven Transformer with LGN-inspired routing.

cs.NE↗

arXiv2026

Subsample-Based Estimation under Dynamic Contamination

This paper studies a structural failure of subsample-based estimation in dynamic time series models. Even under oracle knowledge of contamination locations, removing contaminated observations does not restore the uncontaminated objective. In such settings, contamination propagates through the residual filter and distorts the estimation criterion. As a result, subsample-based estimators are generically inconsistent for the clean-data parameter. We characterise this failure as a structural incompatibility between pointwise subsampling and residual propagation. More generally, the failure arises whenever contamination propagates through transformations that enter the estimation criterion, with dynamic time series models as a leading example. To address it, we propose a propagation-compatible transformation of index sets via a patch removal operator. Under general high-level conditions, this transformation leaves the estimator asymptotically unchanged under the uncontaminated model while restoring consistency under contamination. The results apply to a broad class of residual-based estimators and do not rely on modelling the contamination process.

stat.ME↗

arXiv2025

Iterative GRPO: Batch-Online Multi-Turn RL via Single-Turn RLHF

Practical LLM agents often operate over multi-turn conversations where success is determined only after the full interaction ends. Most multi-turn RL methods train via on-policy rollouts, but unlike in single-turn RLHF, the policy cannot produce a trajectory alone, since an external environment must respond after each agent turn. For conversational agents, this environment is a user, but real users are generally unavailable inside the training loop and simulated users are difficult to build faithfully. Separately, real-world deployment is rarely fully online or fully offline. The common production pattern is called "batch online," where the current policy is deployed to collect a batch of interaction data, then retrained on that batch and redeployed. We show that this batch-online setting creates an opportunity for applying the classical approximate policy iteration algorithm. Our central observation is that running standard token-level, single-turn GRPO with a learned turn-level $Q$-function as the reward model is a policy improvement step for the multi-turn problem. Building on this, we present Iterative GRPO, which alternates between fitting $Q^\pi$ from logged Monte Carlo returns via standard reward modeling (policy evaluation) and running single-turn GRPO against $Q^\pi$ (policy improvement). During policy improvement, the learned $Q^\pi$ scores candidate responses by their expected downstream return, avoiding the need for simulator rollouts of the remaining conversation. Overall, this allows us to do principled multi-turn RL using only single-turn RLHF methods without an interactive environment or user simulator inside the training loop. We demonstrate the effectiveness of Iterative GRPO on six multi-turn negotiation environments.

cs.LG↗

arXiv2025

Registering the 4D Millimeter Wave Radar Point Clouds Via Generalized Method of Moments

4D millimeter wave radars (4D radars) are new emerging sensors that provide point clouds of objects with both position and radial velocity measurements. Compared to LiDARs, they are more affordable and reliable sensors for robots' perception under extreme weather conditions. On the other hand, point cloud registration is an essential perception module that provides robot's pose feedback information in applications such as Simultaneous Localization and Mapping (SLAM). Nevertheless, the 4D radar point clouds are sparse and noisy compared to those of LiDAR, and hence we shall confront great challenges in registering the radar point clouds. To address this issue, we propose a point cloud registration framework for 4D radars based on Generalized Method of Moments. The method does not require explicit point-to-point correspondences between the source and target point clouds, which is difficult to compute for sparse 4D radar point clouds. Moreover, we show the consistency of the proposed method. Experiments on both synthetic and real-world datasets show that our approach achieves higher accuracy and robustness than benchmarks, and the accuracy is even comparable to LiDAR-based frameworks.

cs.RO↗

arXiv2025

Design-Based Inference under Random Potential Outcomes

We study whether mechanism-level causal estimands, defined as expectations over latent stochastic environments, can be consistently recovered from a single realised randomised experiment. Identification alone does not guarantee recoverability. The target estimand averages over latent environments, whereas a single experiment provides only one realisation of such an environment. We show that suitably sparse local dependence induces an ergodic-type property under which cross-sectional averaging consistently recovers expectations over the latent outcome-generating mechanism. Under this structure, aggregate design-based estimators are consistent and asymptotically normal, and the variance becomes consistently estimable from a single experiment. Unlike classical finite-population inference, where Neyman-type variance estimators are structurally limited to conservative upper bounds, the proposed framework permits consistent variance estimation through the shift from fixed potential outcome schedules to stochastic mechanisms.

stat.ME↗

arXiv2021

Elephant-Human Conflict Mitigation: An Autonomous UAV Approach

Elephant-human conflict (EHC) is one of the major problems in most African and Asian countries. As humans overutilize natural resources for their development, elephants' living area continues to decrease; this leads elephants to invade the human living area and raid crops more frequently, costing millions of dollars annually. To mitigate EHC, in this paper, we propose an original solution that comprises of three parts: a compact custom low-power GPS tag that is installed on the elephants, a receiver stationed in the human living area that detects the elephants' presence near a farm, and an autonomous unmanned aerial vehicle (UAV) system that tracks and herds the elephants away from the farms. By utilizing proportional-integral-derivative controller and machine learning algorithms, we obtain accurate tracking trajectories at a real-time processing speed of 32 FPS. Our proposed autonomous system can save over 68 % cost compared with human-controlled UAVs in mitigating EHC.

cs.MA↗

arXiv2019

A Flexible Mixed-Frequency Vector Autoregression with a Steady-State Prior

We propose a Bayesian vector autoregressive (VAR) model for mixed-frequency data. Our model is based on the mean-adjusted parametrization of the VAR and allows for an explicit prior on the 'steady states' (unconditional means) of the included variables. Based on recent developments in the literature, we discuss extensions of the model that improve the flexibility of the modeling approach. These extensions include a hierarchical shrinkage prior for the steady-state parameters, and the use of stochastic volatility to model heteroskedasticity. We put the proposed model to use in a forecast evaluation using US data consisting of 10 monthly and 3 quarterly variables. The results show that the predictive ability typically benefits from using mixed-frequency data, and that improvements can be obtained for both monthly and quarterly variables. We also find that the steady-state prior generally enhances the accuracy of the forecasts, and that accounting for heteroskedasticity by means of stochastic volatility usually provides additional improvements, although not for all variables.

econ.EM↗