Source author record

Xinglin Wang

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Computation and Language Artificial Intelligence cond-mat.soft physics.chem-ph

Catalog footprint

Research overview

3works

4topics

4close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2026arXiv

On Time, Within Budget: Constraint-Driven Online Resource Allocation for Agentic Workflows

Agentic systems increasingly solve complex user requests by executing orchestrated workflows, where subtasks are assigned to specialized models or tools and coordinated according to their dependencies. While recent work improves agent efficiency by optimizing the performance--cost--latency frontier, real deployments often impose concrete requirements: a workflow must be completed within a specified budget and before a specified deadline. This shifts the goal from average efficiency optimization to maximizing the probability that the entire workflow completes successfully under explicit budget and deadline constraints. We study \emph{constraint-driven online resource allocation for agentic workflows}. Given a dependency-structured workflow and estimates of success rates and generation lengths for each subtask--model pair, the executor dynamically allocates models and parallel samples across simultaneously executable subtasks while managing the remaining budget and time. We formulate this setting as a finite-horizon stochastic online allocation problem and propose \emph{Monte Carlo Portfolio Planning} (MCPP), a lightweight closed-loop planner that directly estimates constrained completion probability through simulated workflow executions and replans after observed outcomes. Experiments on CodeFlow and ProofFlow demonstrate that MCPP consistently improves constrained completion probability over strong baselines across a wide range of budget--deadline constraints.

preprint2023arXiv

BatchEval: Towards Human-like Text Evaluation

Significant progress has been made in automatic text evaluation with the introduction of large language models (LLMs) as evaluators. However, current sample-wise evaluation paradigm suffers from the following issues: (1) Sensitive to prompt design; (2) Poor resistance to noise; (3) Inferior ensemble performance with static reference. Inspired by the fact that humans treat both criterion definition and inter sample comparison as references for evaluation, we propose BatchEval, a paradigm that conducts batch-wise evaluation iteratively to alleviate the above problems. We explore variants under this paradigm and confirm the optimal settings are two stage procedure with heterogeneous batch composition strategy and decimal scoring format. Comprehensive experiments across 3 LLMs on 4 text evaluation tasks demonstrate that BatchEval outperforms state-of-the-art methods by 10.5% on Pearson correlations with only 64% API cost on average. Further analyses have been conducted to verify the robustness, generalization, and working mechanism of BatchEval.

preprint2020arXiv

Elucidating the $^1$ H NMR relaxation mechanism in polydisperse polymers and bitumen using measurements, MD simulations, and models

The mechanism behind the $^1$ H NMR frequency dependence of $T_1$ and the viscosity dependence of $T_2$ for polydisperse polymers and bitumen remains elusive. We elucidate the matter through NMR relaxation measurements of polydisperse polymers over an extended range of frequencies ( $f_0 = 0.01 \leftrightarrow$ 400 MHz) and viscosities ( $η= 385 \leftrightarrow 102,000$ cP) using $T_{1}$ and $T_2$ in static fields, $T_{1}$ field-cycling relaxometry, and $T_{1ρ}$ in the rotating frame. We account for the anomalous behavior of the log-mean relaxation times $T_{1LM} \propto f_0$ and $T_{2LM} \propto (η/T)^{-1/2}$ with a phenomenological model of $^1$ H- $^1$ H dipole-dipole relaxation which includes a distribution in molecular correlation times and internal motions of the non-rigid polymer branches. We show that the model also accounts for the anomalous $T_{1LM}$ and $T_{2LM}$ in previously reported bitumen measurements. We find that molecular dynamics (MD) simulations of the $T_{1} \propto f_0$ dispersion and $T_2$ of similar polymers simulated over a range of viscosities ( $η= 1 \leftrightarrow 1,000$ cP) are in good agreement with measurements and the model. The $T_{1} \propto f_0$ dispersion at high viscosities agrees with previously reported MD simulations of heptane confined in a polymer matrix, which suggests a common NMR relaxation mechanism between viscous polydisperse fluids and fluids under confinement, without the need to invoke paramagnetism.

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Xinglin Wang

Research overview

Is this your work?

3 publications

On Time, Within Budget: Constraint-Driven Online Resource Allocation for Agentic Workflows

BatchEval: Towards Human-like Text Evaluation

Elucidating the $^1$ H NMR relaxation mechanism in polydisperse polymers and bitumen using measurements, MD simulations, and models

See the researcher in context

Building this map preview

Xinglin Wang

Research overview

Is this your work?

3 publications

On Time, Within Budget: Constraint-Driven Online Resource Allocation for Agentic Workflows

BatchEval: Towards Human-like Text Evaluation

Elucidating the 1^11H NMR relaxation mechanism in polydisperse polymers and bitumen using measurements, MD simulations, and models

See the researcher in context

Building this map preview

Elucidating the $^1$ H NMR relaxation mechanism in polydisperse polymers and bitumen using measurements, MD simulations, and models