Searcharxiv⌕ Search

arXiv subjects

Yu Cao

Publications and source records attributed to Yu Cao.

At least 19 recordsLinked to original sources
arXiv2026

Counterexamples to the Tensorization Property of Strong Data Processing Inequalities for Quantum Divergences

In information theory, the data processing inequality is a fundamental property that describes the loss of information through noisy channels. A more refined description is given by the strong data processing inequality (SDPI) for a fixed channel and a full-rank reference state. In classical information theory, the strong data processing inequality tensorizes for a large family of $f$-divergences. However, its quantum counterpart remains largely unexplored. Tensorization of SDPI has previously been shown only for some special cases, and whether it holds more generally for quantum divergences still remains open. In this work, we report two negative results: the tensorization property fails for certain quantum chi-square divergences, and it also does not hold for the quantum relative entropy.

quant-ph↗
arXiv2026

Auditing evaluation artifacts and data leakage in large language model systems for rare disease diagnosis

Augmented large language model (LLM) systems have reported high benchmark accuracies in rare disease diagnosis, yet their real-world clinical utility remains unproven. A recent meta-analysis flagged augmented LLM diagnostic evaluations as having a high risk of bias due to suspected data leakage; however, the precise impact of these vulnerabilities on model inference has yet to be fully quantified. In this study, we present a comprehensive empirical audit to quantify evaluation artifacts across prevailing AI paradigms, using a repository of over 60,000 cases from diverse public and independent clinical cohorts. By reconstructing a state-of-the-art agentic framework, we demonstrate that candidate-space contamination artificially inflates performance: enforcing strict cross-dataset isolation resulted in a marked performance decline, with the cross-dataset mean Recall@1 decreasing from 54.3% (under conventional one-vs-rest evaluation) to 40.3% (Hedges' g = 0.94). Similar spatial overfitting in supervised classifiers caused diagnostic accuracy to drop from an internal mean of 42.9% to 16.7% across external cohorts (Cohen's d = 2.01). Furthermore, we quantify how LLMs infer diagnoses by exploiting retrospective therapeutic cues in clinical narratives, and how temporal misalignment in foundational knowledge bases leads to performance fluctuations. We conclude that prevailing evaluation protocols for augmented LLM systems inadvertently reflect data leakage rather than genuine medical reasoning. To address these systematic artifacts, we propose a rigorous evaluation framework - coupling strict data-isolation protocols with standardized benchmarking safeguards - to ensure future diagnostic AI for rare diseases is evaluated on actual clinical robustness.

cs.AI↗
arXiv2026

Bridging the Data Gap: Source-Informed Task-Space Transport for Few-Shot Diffusion

Many deployments require a model to move rapidly from a data-rich source to a related target observed only sparsely. This is especially consequential where target data are difficult to collect, as in healthcare settings with persistent population or acquisition imbalances. Retraining or fine-tuning a high-capacity generator for every target is costly and can be statistically unreliable when target data are scarce. We introduce Meta-Diff, which instead learns a recurring source--target relation once in a compact task space. A shared encoder maps each distribution to a low-dimensional coordinate; a learned relation transports a new source coordinate to its predicted target, and a shared diffusion backbone generates from that prediction without target-specific retraining. Controlled Gaussian-mixture experiments identify when this works: task identity must survive the representation, the relation must generalise across tasks, and the source--target shift must leave useful headroom for transport. On held-out CIFAR-100 validation tasks, correct source $+$ transport achieves 13.9% lower mean SW than transport from the wrong source. With validation-selected adaptation settings in the final-test one-target comparison, its mean SW is 14.7% lower than target-only modelling and 32.7% lower than full fine-tuning; the main route itself uses no target image and assumes the applicable relation is known. On matched CIFAR test episodes, generating from the clean-source coordinate and then applying the known corruption yields lower SW than task-space transport. Office-Home exposes a complementary boundary: transport can align target-domain statistics without reliable object identity. Together, these results recast extreme-few-shot adaptation as reusable transport in task space, while making the conditions for its success and failure directly measurable.

cs.LG↗
arXiv2026

Unified Multiparadigm Fusion Intelligence for Rare Disease Diagnosis

Rare diseases (RDs) affect an estimated 300 million people globally, leaving patients to navigate a protracted diagnostic odyssey constrained by clinical scarcity and fragmented medical knowledge. Artificial intelligence (AI) has progressively advanced RD diagnostics, from symbolic methods and deep learning to large language model (LLM)-based systems. However, each paradigm carries structural limitations: symbolic methods are bounded by knowledge completeness, neural models sacrifice interpretability, and LLM-based systems are vulnerable to evaluation biases like data leakage. In this study, we introduce FusionRare, a unified multiparadigm fusion framework integrating knowledge-driven symbolic reasoning, deep-learning pattern recognition, case-based patient retrieval, and LLM-based heuristic inference via statistical meta-aggregation and knowledge-enhanced agentic reflection. Evaluated on 2,528 strictly isolated clinical cases, FusionRare achieves state-of-the-art RD prioritization, attaining a macro-averaged Recall @1 of 51% for phenotype-only diagnosis and 79% for phenotype-genotype joint analysis. These results represent decisive advantages over the advanced agentic system DeepRare, frontier LLMs including GPT 5.5, and specialized tools like PhenoBrain and LIRICAL under stringent anti-leakage protocols. A randomized, crossover clinician-AI validation demonstrated that FusionRare elevates the true diagnostic rank in 45% of decisions, improves physician confidence in 62%, and reduces time to diagnosis by 21%. These findings indicate that reconciling these orthogonal AI paradigms significantly enhances diagnostic capabilities, bridging the gap between computational benchmarks and actionable clinical reliability. More broadly, this work provides a replicable framework for deploying trustworthy diagnostic AI across complex, high-stakes, and underserved medical domains.

cs.AI↗
arXiv2026

RELAY: Selective Self-Distillation with Adaptive Hand-off for Rubric-Based Reinforcement Learning

Rubric-based reinforcement learning grades a response against many written criteria but collapses them into one reward, which discards the detail the rubric carries and gives no GRPO reward gradient when a group's responses tie. A frozen teacher that reads the rubric as privileged context provides token-level supervision, but we observe an early plateau in the gains over GRPO. RELAY uses advantages and rewards to select positive-advantage and low-scoring zero-advantage rollouts, and monitors the distillation loss to decide when to retire the teacher. Stage 1 combines selective distillation from a frozen teacher with GRPO; a plateau detector then removes the teacher, and Stage 2 continues with GRPO alone. Calibrated on one model-domain pair and reused unchanged in three others, the plateau detector fires between steps 122 and 171. RELAY scores highest in-domain among the compared methods in all four settings, ahead of GRPO on RaR-Medicine by 4.62 points with Qwen3-1.7B and 6.02 with Qwen3-4B. On Qwen3-1.7B/Medicine, removing the teacher from a shared checkpoint outperforms keeping, weakening or refreshing it. Distilling selected rollouts covering about half the response tokens also outperforms distilling all rollouts at a matched average coefficient.

cs.LG↗
arXiv2026

Bridging the Last Mile of Circuit Design: PostEDA-Bench, a Hierarchical Benchmark for PPA Convergence and DRC Fixing

LLM-based agents are increasingly applied to the "last mile" of Electronic Design Automation (EDA): repairing residual sign-off Design Rule Check (DRC) violations and converging Power-Performance-Area (PPA) targets after tool runs. Existing EDA-LLM benchmarks, however, omit DRC fixing entirely and rely on flat hierarchies tied to a single toolchain. We introduce PostEDA-Bench, a hierarchical benchmark with 145 tasks across DRC-Essential, DRC-Reasoning, PPA-Mono, and PPA-Multi, supported by EDA toolchains with machine-checkable evaluation. Across eight commercial and open-source LLMs under multiple agent scaffolds, we find that agents handle synthetic DRC-Essential and single-objective PPA-Mono reasonably well but degrade sharply on the more practical DRC-Reasoning, where the best success rate is 36.66%, and PPA-Multi, where the best success rate is 20.00%; vision augmentation consistently enhances DRC-Bench; and trade-off reasoning, rather than knob knowledge, is the dominant PPA-Multi bottleneck.

cs.AR↗
arXiv2026

Reflex-Informed Neuromuscular Reinforcement Learning for Muscle-Driven Locomotion

Muscle-driven locomotion provides a physically grounded approach to generating realistic human movement. However, achieving both physiological plausibility and adaptability to changes in musculoskeletal capacity and external disturbances remains a fundamental challenge. To address this limitation, we propose a Reflex-Informed Neuromuscular Reinforcement Learning framework for muscle-driven locomotion. Within this framework, a fixed phase-dependent reflex controller serves as the underlying neuromuscular control mechanism, while the reinforcement learning policy produces four biomechanically meaningful residual parameters to modulate key reflex gains and thresholds associated with hip swing, knee support, and ankle propulsion according to the current state. Experimental results demonstrate that the proposed framework generates physiologically plausible locomotion with improved kinematic accuracy and dynamic consistency, as well as better bilateral symmetry and stride-to-stride consistency under nominal walking conditions. The learned policy remains robust under muscle weakness and external perturbations without retraining.

cs.RO↗
arXiv2026

LACE: Large Language Model Aided Multi-Agent Framework for Agile RISC-V Instruction Extension

Domain-specific Instruction Set Architecture eXtensions (ISAX) are widely adopted in the RISC-V ecosystem to accelerate emerging workloads, but implementing and validating ISAXes across different cores remains slow and fragmented. Existing frameworks still require per-core interface adaptation, and differential testing often breaks once either the microarchitecture or the ISAX changes. We present LACE, an LLM-aided multi-agent workflow that translates natural-language ISAX intents into a compact two-level IR (operation-level and HDL task-level), performs retrieval-guided localized RTL edits over large repositories, and closes the loop with a compiler-agnostic riscv-formal checking flow (assuming RVFI availability or instrumentation). Across four embedded RISC-V cores, LACE raises pass@1 generation accuracy from near-zero to 72.8\% within our evaluation setup, while improving code localization and reducing integration rework. The code of LACE is available at https://github.com/UMN-ZhaoLab/LACE.

cs.AR↗
arXiv2026

SERPO: Self-Evolving Rubric Policy Optimization for Open-Ended Test-Time Reinforcement Learning

Test-time reinforcement learning (TTRL) enables language models to self-evolve at inference time without labeled feedback. Existing methods rely on answer voting and therefore do not extend naturally to open-ended generation, where valid responses cannot be mapped to a shared canonical answer. Without external reward models or stronger judges, adaptation must instead construct reliable rewards from the model's own outputs. We introduce SERPO (Self-Evolving Rubric Policy Optimization), which replaces answer voting with a closed loop that co-evolves response evidence, query-specific rubrics, and policy parameters. Good-Normal-Bad (G-N-B) response evolution organizes maximally separated rollouts into ordered archives; rubric evolution retains criteria that discriminate these archives; probabilistic criterion scoring converts verdict-token likelihoods into reward signals; and policy evolution optimizes the actor with the resulting signals. New actor rollouts then refresh both the archives and rubrics, closing the three-way evolution loop. Across two model configurations, two in-domain benchmarks, and four OOD benchmarks, SERPO improves HealthBench and ResearchQA by up to 20.63 and 20.31 points over the corresponding base models, raises the six-benchmark macro-average by up to 8.06 points, and supports OOD transfer and continued cross-benchmark evolution.

cs.CL↗
arXiv2026

From Papers to Interpretive Knowledge Nodes: Proposing the Missing Object in Scholarly Knowledge Circulation

The modern scholarly communication system, with the paper at its core, has successfully solidified "research outputs" into citable and traceable scholarly objects. Yet, across the full chain from knowledge production to knowledge reuse, a critical link - interpretation - has long existed without ever being objectified. The theoretical elaborations, methodological translations, and conceptual clarifications that researchers perform when reading papers constitute the factual foundation of knowledge circulation. However, these interpretive activities remain tethered to individual competence and informal communication channels, lacking an independent scholarly identity. The central question of this paper is: Can interpretive knowledge qualify as an independent scholarly object? We argue that the Interpretive Knowledge Node (IKN), while manifesting as a "node" within knowledge networks, should be defined in its object identity as a new kind of scholarly object. Drawing on scholarly object theory, this paper identifies a structural gap in the existing system - namely, that "interpretive knowledge formed after the process of understanding" lacks an objectified identity. By distinguishing the essential difference between "interpretation" and "information extraction," it establishes the cognitive boundary of the IKN. Finally, it defines the four essential attributes that an IKN must possess as a scholarly object: provenance, interpretation fixation, human recognition, and citability. The argument demonstrates that the IKN is not a replacement for the paper system but an expansion of the boundary of the scholarly object system - marking a critical step from merely recording "what has been discovered" toward simultaneously recording "how to understand those discoveries."

cs.DL↗
arXiv2026

From Compressing Complexity to Accommodating Complexity: How AI Transforms Standardization and Individualization

Why do societies composed of individuals pursuing individuality repeatedly generate highly standardized systems? This paper argues that the answer lies in the evolution of information processing capacity. Artificial intelligence represents a historical transition in this capacity, enabling social systems to accommodate forms of complexity that previously had to be compressed. Industrial standardization was not merely a consequence of capital preference or power relations, but an institutional arrangement for maintaining the manageability of large-scale systems under limited information-processing capacity by reducing the variety of the controlled system. The fundamental change in the AI era lies in the expansion of information processing capacity across three dimensions: perception, computation, and execution. This expansion shifts personalized production from physical adaptation toward information-based adaptation and enables a transition from discrete to continuous objectification of difference. This paper proposes "cognitive fixed cost" as an analytical concept to describe how the upfront concentration of cognitive labor transforms the cost structure of personalized production. It further argues that standardization has not disappeared but has moved from explicit constraints at the product level to implicit generation rules embedded in infrastructures, shifting the central contradiction from "whether to have commonality" to "who controls commonality." The evolution of civilizational production logic is not a movement from commonality to individuality, but from compressing complexity to accommodating complexity.

cs.CY↗
arXiv2026

Citation Pathways in the AI Era: Interpretive Knowledge Nodes, Citation Compression Layers, and the Measurement Boundary of Scholarly Impact

This article identifies "citation pathway" as a long-neglected analytical dimension in scientometrics. Traditional evaluation metrics focus on measuring citation counts while paying insufficient attention to the intermediate nodes through which knowledge flows from its original source to the citing author. Building on an analysis of the normative structure of current reference systems, this article introduces two new concepts: Interpretive Knowledge Nodes (IKN) - academic papers that provide structured reorganizations of classic works - and Citation Compression Layers (CCL) - the intermediate layers that emerge when such knowledge products acquire stable publication identities and enter formal citation networks at scale. The central proposition is that AI has not changed citation rules themselves but has transformed the cost structure of producing citable knowledge intermediaries. Under conditions of full compliance, the network position effects of citation pathways may become a salient variable affecting the validity of impact measurement. Through a thought experiment involving a hypothetical journal R and a parsimonious "Citation Gravity conceptual model," this article substantiates this proposition and discusses the measurement boundaries of scholarly impact indicators, as well as the institutional risk of incentive misalignment under extreme scenarios.

physics.soc-ph↗
arXiv2026

Correlation-consistent Gaussian basis sets for copper solids from material-constrained atomic optimization

Correlation-consistent Gaussian basis sets are central to systematic molecular quantum chemistry, but their direct use in periodic solids is often limited by severe linear dependence from diffuse atomically optimized primitives. This problem is particularly acute for metallic and metal-containing systems, where reliable complete-basis-set (CBS) extrapolation is needed for correlated-wavefunction benchmarks. We introduce material-constrained atomic optimization (MCAO), a basis-set optimization framework that preserves the atomic and correlation-consistent character of Gaussian basis sets while penalizing large overlap-matrix condition numbers in representative solids. As a proof of concept, we generate Dunning-style MCAO-cc-pVXZ basis sets (X = D, T, Q) for Cu with all-electron, scalar-relativistic all-electron, effective core potential (ECP), and pseudopotential treatments. The resulting basis sets remain numerically stable for Cu solids and surfaces while reproducing molecular Cu dimer energetics and plane-wave reference properties of bulk Cu. CBS-extrapolated random-phase approximation calculations further enable a controlled assessment of pseudopotential, relativistic, and basis-set errors in bulk Cu and CO adsorption on Cu(111), providing scalar-relativistic all-electron Gaussian-basis benchmarks for the CO adsorption puzzle.

physics.chem-ph↗
arXiv2026

Grade Encoding and the Structural Representation of Student Academic Trajectories

Does the conversion of academic assessment from percentage scores to letter grades merely represent an adjustment in information precision, or does it systematically alter the underlying structure of student academic data? Drawing on 68 mathematics exam scores from 75 primary school students, this study employs Encoding Transformation simulation to compare structural differences in the same dataset under two encoding schemes across three dimensions: information loss, distance structure change, and clustering stability. Results indicate that letter-grade encoding compresses the mean pairwise distance in the trajectory feature space from 20.50 to 1.06 (a compression ratio of approximately 19:1); after standardization, the density gradient of the distance distribution is systematically flattened, with kurtosis decreasing by 0.54 and the coefficient of variation decreasing by 0.16; and the clustering structure becomes highly sensitive to minor sample perturbations-removing a single extreme observation causes the optimal K to jump from 4 to 8 and clustering reproducibility to plummet from 95% to 62%. Concurrently, a Compression Paradox is observed: letter-grade encoding increases the Silhouette coefficient while simultaneously reducing clustering stability. These findings demonstrate that letter-grade encoding systematically alters the structural representation of longitudinal academic data and exerts measurable effects on clustering stability.

cs.CY↗
arXiv2026

Efficient Bethe-Salpeter Equation Calculations Based on Numerical Atomic Orbitals and Norm-Conserving Pseudopotentials: Dual-${\boldsymbol k}$-Mesh Strategy

We present an efficient implementation of the Bethe--Salpeter equation (BSE) based on numerical atomic orbitals (NAOs) and norm-conserving pseudopotentials within the ABACUS+LibRPA framework. By exploiting the localized resolution-of-identity (LRI) technique, the screened Coulomb interaction is cast into a real-space, unit-cell-indexed form $W_{μν}(\boldsymbol R)$ that is inherently short-ranged and well localized. This spatial locality enables an efficient Fourier interpolation of the BSE kernel from the coarse $\boldsymbol k$-mesh used in the preceding $GW$ calculation to an arbitrarily dense $\boldsymbol k$-mesh on which the BSE Hamiltonian is assembled and diagonalized, thereby giving rise naturally to a dual-$\boldsymbol k$-mesh workflow. Building on this scheme, we systematically examine the convergence of the absorption spectra with respect to the NAO basis set, the auxiliary basis set, and the $\boldsymbol k$-point sampling. Benchmark calculations for both molecular and periodic systems collectively validate the accuracy of the present implementation and establish the dual-$\boldsymbol k$-mesh strategy as a practical and reliable approach for $GW$+BSE calculations.

cond-mat.mtrl-sci↗
arXiv2026

LatSearch: Latent Reward-Guided Search for Faster Inference-Time Scaling in Video Diffusion

The recent success of inference-time scaling in large language models has inspired similar explorations in video diffusion. In particular, motivated by the existence of "golden noise" that enhances video quality, prior work has attempted to improve inference by optimising or searching for better initial noise. However, these approaches have notable limitations: they either rely on priors imposed at the beginning of noise sampling or on rewards evaluated only on the denoised and decoded videos. This leads to error accumulation, delayed and sparse reward signals, and prohibitive computational cost, which prevents the use of stronger search algorithms. Crucially, stronger search algorithms are precisely what could unlock substantial gains in controllability, sample efficiency and generation quality for video diffusion, provided their computational cost can be reduced. To fill in this gap, we enable efficient inference-time scaling for video diffusion through latent reward guidance, which provides intermediate, informative and efficient feedback along the denoising trajectory. We introduce a latent reward model that scores partially denoised latents at arbitrary timesteps with respect to visual quality, motion quality, and text alignment. Building on this model, we propose LatSearch, a novel inference-time search mechanism that performs Reward-Guided Resampling and Pruning (RGRP). In the resampling stage, candidates are sampled according to reward-normalised probabilities to reduce over-reliance on the reward model. In the pruning stage, applied at the final scheduled step, only the candidate with the highest cumulative reward is retained, improving both quality and efficiency. We evaluate LatSearch on the VBench-2.0 benchmark and demonstrate that it consistently improves video generation across multiple evaluation dimensions compared to the baseline Wan2.1 model.

cs.CV↗
arXiv2026

Each Judge Its Own Yardstick: Discovering Per-VLM Taxonomies for Physical Video Evaluation

Maintaining physical consistency in video generators and world models increasingly relies on vision-language models (VLMs) as automated judges that provide reward signals, ranking decisions, and data-filtering criteria. Yet VLMs differ substantially in training data and architecture, encoding physical phenomena through distinct internal representations. A single global evaluation schema therefore gives every VLM the same axes of competence, regardless of what each can actually perceive. We propose JudgeFit, an iterative refinement procedure that discovers a per-VLM evaluation taxonomy. An initial taxonomy is constructed by prompting the target VLM to enumerate physics errors on a small set of videos and clustering the resulting descriptions. The taxonomy is then refined through a diagnostic step: we calibrate the VLM's per-dimension scores to human physical-commonsense ratings, diagnose which dimensions it scores unreliably or redundantly, and prompt an LLM to repair them, iterating until convergence. We further instantiate this procedure as a benchmark and apply it to 16 VLMs spanning eight model families. The refined taxonomy outperforms the global-schema baseline on held-out videos for every VLM tested, with a mean relative improvement of approximately 32%. Beyond aggregate accuracy, the per-VLM profiles expose model-specific blind spots that overall rankings cannot anticipate, with reliability patterns differing markedly across model families.

cs.CV↗
arXiv2026

Cross-Stage Sensorimotor Perception Scheduling and Sparse Map Encoding for Efficient Edge Embodied Navigation

Embodied agents must close a perception-to-action loop on embedded hardware under tight latency, memory, and energy budgets, making deployment a system-level co-design problem rather than a model-accuracy problem. We study this challenge for modular Object Goal Navigation (ObjectNav), where our profiling shows semantic mapping dominates per-step latency while goal prediction dominates peak memory. We formulate edge embodied navigation deployment as a budget-constrained design-space problem and introduce two orthogonal optimization knobs: SKIP, an adaptive sensorimotor scheduler that formalizes safe skipping as a bounded map-impact criterion and learns a lightweight predictor to estimate it from cheap sensor cues at each \texttt{FORWARD} step, exposing a principled quality-efficiency knob (depth-based updates are always retained); and SCOUT, a sparse-context encoder that couples submanifold sparse convolutions on active map regions with a lightweight dense context stream. On HM3D across server and embedded platforms, SKIP+SCOUT delivers up to 1.7x end-to-end speedup, 50.5% lower peak memory, and 7.1% higher SPL than the dense baseline at the selected operating point, outperforming naively smaller perception backbones. SKIP transfers to a second modular pipeline (PONI) with near-lossless performance and remains robust under depth-sensor noise. Together, SKIP+SCOUT expose a family of device-aware Pareto operating points for edge physical AI systems.

cs.RO↗