arXiv subjects

Shuyu Liu

Publications and source records attributed to Shuyu Liu.

9 recordsLinked to original sources

arXiv2026

What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics

Task completion is the standard metric for evaluating context compression, yet it is incomplete: compression can increase an agent's interaction cost by forcing it to reacquire dropped state while leaving completion statistically unchanged. We introduce a controlled runtime measurement protocol for reacquisition cost in a bounded-horizon tool-using agent. The agent acts in a deterministic planning environment under a fixed 24-turn horizon. We vary compression severity, compare a dropping operator with a fact-preserving operator, restore dropped state through controlled oracle interventions, and decompose tool calls into retrieval and execution. We evaluate three models across two task regimes. Retrieval calls increase in all six model-regime comparisons and account for almost all added interaction; five of six remain significant after Holm correction. At the prespecified 5x comparison point, completion changes are not significant in any cell. DeepSeek shows a significant completion drop only at 10x compression. GPT-5.5 is the clearest case: completion changes from 80% to 85% (p = 1.0) while retrieval increases from 21.0 to 63.9 calls (p = .002). Retention interventions further separate state quantity, state type, and content validity. Random selection is comparable to an offline hindsight oracle, while replacing retained D-state with semantically irrelevant content increases retrieval by 57% (p < .001) without a significant completion change. In a second environment, ALFWorld, sliding compression produces no retrieval surge, showing that the reacquisition signature is environment-dependent rather than intrinsic to shortening context. Overall, compression can impose hidden interaction costs when execution-relevant state becomes absent and must be reacquired, while completion alone may not expose those costs.

cs.AI↗

arXiv2026

Mobile Target Search with Imperfect Perception: A Partially Observable Stochastic Game Theoretical Approach

This paper investigates mobile target search under imperfect perceptions caused by sensor limitations, malicious jamming, or communication noise. Searchers and targets operate in a grid-shaped area with bounded mobility, leading to a dynamic interplay between search and evasion. To capture this adversarial interaction under imperfect perceptions, we adopt the partially observable stochastic game (POSG) approach, which generalizes partially observable Markov decision processes (POMDPs) by incorporating target intelligence. To handle false alarms and missed detections caused by perceptual uncertainties, we propose a novel detectability concept to determine whether a search strategy guarantees eventual detection, and provide sufficient detectability criteria based on stochastic recurrence analysis. We further develop a server-assisted distributed algorithm that utilizes the aggregative potential game structure for searchers and a KL-divergence-based reduction for target prediction. Numerical simulations validate the effectiveness of the proposed algorithm and support the detectability analysis.

cs.RO↗

arXiv2026

Generative Inverse Design of Cold Metals for Low-Power Electronics

Cold metals are a class of metals with an intrinsic energy gap located close to the Fermi level, which enables cold-carrier injection for steep-slope transistors and is therefore promising for low-power electronic applications. High-throughput screening has revealed 252 three-dimensional (3D) cold metals in the Materials Project database, but database searches are inherently limited to known compounds. Here we present an inverse-design workflow that generates 3D cold metals using MatterGPT, a conditional autoregressive Transformer trained on SLICES, an invertible and symmetry-invariant crystal string representation. We curate a training set of 26,309 metallic structures labeled with energy above hull and a unified band-edge distance descriptor that merges p-type and n-type cold-metal characteristics to address severe label imbalance. Property-conditioned generation targeting thermodynamic stability and 50-500 meV band-edge distances produces 148,506 unique candidates; 92.1% are successfully reconstructed to 3D structures and down-selected by symmetry, uniqueness and novelty filters, followed by high-throughput DFT validation. We identify 257 cold metals verified as novel with respect to the Materials Project database, with gaps around the Fermi level spanning 50-500 meV. First-principles phonon, electronic-structure, and work-function calculations for representative candidates confirm dynamical stability and contact-relevant work functions. Our results demonstrate that SLICES-enabled generative transformers can expand the chemical space of cold metals beyond high-throughput screening, providing a route to low-power electronic materials discovery.

cond-mat.mtrl-sci↗

arXiv2025

PsychBench: A comprehensive and professional benchmark for evaluating the performance of LLM-assisted psychiatric clinical practice

The advent of Large Language Models (LLMs) offers potential solutions to address problems such as shortage of medical resources and low diagnostic consistency in psychiatric clinical practice. Despite this potential, a robust and comprehensive benchmarking framework to assess the efficacy of LLMs in authentic psychiatric clinical environments is absent. This has impeded the advancement of specialized LLMs tailored to psychiatric applications. In response to this gap, by incorporating clinical demands in psychiatry and clinical data, we proposed a benchmarking system, PsychBench, to evaluate the practical performance of LLMs in psychiatric clinical settings. We conducted a comprehensive quantitative evaluation of 16 LLMs using PsychBench, and investigated the impact of prompt design, chain-of-thought reasoning, input text length, and domain-specific knowledge fine-tuning on model performance. Through detailed error analysis, we identified strengths and potential limitations of the existing models and suggested directions for improvement. Subsequently, a clinical reader study involving 60 psychiatrists of varying seniority was conducted to further explore the practical benefits of existing LLMs as supportive tools for psychiatrists of varying seniority. Through the quantitative and reader evaluation, we show that while existing models demonstrate significant potential, they are not yet adequate as decision-making tools in psychiatric clinical practice. The reader study further indicates that, as an auxiliary tool, LLM could provide particularly notable support for junior psychiatrists, effectively enhancing their work efficiency and overall clinical quality. To promote research in this area, we will make the dataset and evaluation framework publicly available, with the hope of advancing the application of LLMs in psychiatric clinical settings.

cs.CL↗

arXiv2025

Beyond entropic regularization: Debiased Gaussian estimators for discrete optimal transport and general linear programs

This work proposes new estimators for discrete optimal transport plans that enjoy Gaussian limits centered at the true solution. This behavior stands in stark contrast with the performance of existing estimators, including those based on entropic regularization, which are asymptotically biased and only satisfy a CLT centered at a regularized version of the population-level plan. We develop a new regularization approach based on a different class of penalty functions, which can be viewed as the duals of those previously considered in the literature. The key feature of these penalty schemes it that they give rise to preliminary estimates that are asymptotically linear in the penalization strength. Our final estimator is obtained by constructing an appropriate linear combination of two penalized solutions corresponding to two different tuning parameters so that the bias introduced by the penalization cancels out. Unlike classical debiasing procedures, therefore, our proposal entirely avoids the delicate problem of estimating and then subtracting the estimated bias term. Our proofs, which apply beyond the case of optimal transport, are based on a novel asymptotic analysis of penalization schemes for linear programs. As a corollary of our results, we obtain the consistency of the naive bootstrap for fully data-driven inference on the true optimal solution. Simulation results and two data analyses support strongly the benefits of our approach relative to existing techniques.

math.ST↗

arXiv2025

Quantifying the Reasoning Abilities of LLMs on Real-world Clinical Cases

Recent advancements in reasoning-enhanced large language models (LLMs), such as DeepSeek-R1 and OpenAI-o3, have demonstrated significant progress. However, their application in professional medical contexts remains underexplored, particularly in evaluating the quality of their reasoning processes alongside final outputs. Here, we introduce MedR-Bench, a benchmarking dataset of 1,453 structured patient cases, annotated with reasoning references derived from clinical case reports. Spanning 13 body systems and 10 specialties, it includes both common and rare diseases. To comprehensively evaluate LLM performance, we propose a framework encompassing three critical examination recommendation, diagnostic decision-making, and treatment planning, simulating the entire patient care journey. To assess reasoning quality, we present the Reasoning Evaluator, a novel automated system that objectively scores free-text reasoning responses based on efficiency, actuality, and completeness using dynamic cross-referencing and evidence checks. Using this benchmark, we evaluate five state-of-the-art reasoning LLMs, including DeepSeek-R1, OpenAI-o3-mini, and Gemini-2.0-Flash Thinking, etc. Our results show that current LLMs achieve over 85% accuracy in relatively simple diagnostic tasks when provided with sufficient examination results. However, performance declines in more complex tasks, such as examination recommendation and treatment planning. While reasoning outputs are generally reliable, with factuality scores exceeding 90%, critical reasoning steps are frequently missed. These findings underscore both the progress and limitations of clinical LLMs. Notably, open-source models like DeepSeek-R1 are narrowing the gap with proprietary systems, highlighting their potential to drive accessible and equitable advancements in healthcare.

cs.CL↗

arXiv2025

PediaBench: A Comprehensive Chinese Pediatric Dataset for Benchmarking Large Language Models

The emergence of Large Language Models (LLMs) in the medical domain has stressed a compelling need for standard datasets to evaluate their question-answering (QA) performance. Although there have been several benchmark datasets for medical QA, they either cover common knowledge across different departments or are specific to another department rather than pediatrics. Moreover, some of them are limited to objective questions and do not measure the generation capacity of LLMs. Therefore, they cannot comprehensively assess the QA ability of LLMs in pediatrics. To fill this gap, we construct PediaBench, the first Chinese pediatric dataset for LLM evaluation. Specifically, it contains 4,117 objective questions and 1,632 subjective questions spanning 12 pediatric disease groups. It adopts an integrated scoring criterion based on different difficulty levels to thoroughly assess the proficiency of an LLM in instruction following, knowledge understanding, clinical case analysis, etc. Finally, we validate the effectiveness of PediaBench with extensive experiments on 20 open-source and commercial LLMs. Through an in-depth analysis of experimental results, we offer insights into the ability of LLMs to answer pediatric questions in the Chinese context, highlighting their limitations for further improvements. Our code and data are published at https://github.com/ACMISLab/PediaBench.

cs.CL↗

arXiv2023

Asymptotic confidence sets for random linear programs

Motivated by the statistical analysis of the discrete optimal transport problem, we prove distributional limits for the solutions of linear programs with random constraints. Such limits were first obtained by Klatt, Munk, & Zemel (2022), but their expressions for the limits involve a computationally intractable decomposition of $\mathbb{R}^m$ into a possibly exponential number of convex cones. We give a new expression for the limit in terms of auxiliary linear programs, which can be solved in polynomial time. We also leverage tools from random convex geometry to give distributional limits for the entire set of random optimal solutions, when the optimum is not unique. Finally, we describe a simple, data-driven method to construct asymptotically valid confidence sets in polynomial time.

math.ST↗

arXiv2022

Characterization of Graphene/Ionic Liquid Memristive Devices for Neuromorphic Systems

Flexible and biocompatible memristive devices are particularly attractive for bioelectronic systems due to the interest in improving computing capabilities and the motivation to interface electronics with biological systems including drug delivery, neural interfaces and biosensors. Structures made of more unorthodox, organic material can address different issues due to their characteristics: flexibility, conformability, biocompatibility and simple and low-cost fabrication. It has been observed that gating graphene/ionic liquid (IL) devices leads to the formation of an electrical double layer (a thin layer of ions with a thickness of a few nanometers) at the graphene/IL interface due to the local potential difference which also controls the local conductivity. This structure provides a memristive mechanism based on a dynamic p-n junction formation along the channel. Motivated by this memristive behavior, graphene/IL devices were assembled with the aim of demonstrating memristive behavior and associative learning. This work investigates memristive properties of flexible graphene/ IL devices on polymer substrates. The I-V characteristics of these novel devices and switching mechanism are investigated. Two distinct topologies (single input, single output and double input, single output) of devices are manufactured and tested to mimic conditioning. It is observed that the application of voltage pulse trains of both positive and negative polarities increases the device conductance and allows larger currents to pass after repetitive excitation. This characteristic was exploited to condition devices and emulate associative learning.

cond-mat.mes-hall↗