arXiv subjects

Zheng Zhang

Publications and source records attributed to Zheng Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2026

Choosing Before Acting: Comparative Value Estimation for Long-Horizon Tool-Use Agents

Large language models (LLMs) rely on long-horizon tool invocation sequences for complex tasks, where each invocation can alter the task state and condition subsequent decisions. In long-horizon tool use, final-outcome rewards provide weak credit assignment over long interaction traces. Step-level rewards can offer more targeted feedback, but obtaining reliable step supervision often requires human or LLM judgment, or additional rollouts to estimate the downstream effect of an intermediate decision. In this paper, we argue that effective tool-use agents should estimate the long-horizon value of a possible next tool invocation before executing it. This objective requires comparative supervision over alternative invocations under the same context, while logged trajectories only contain the invocation that was actually taken. Therefore, we propose Comparative Inference for Tool-use Agents (CITA). CITA trains a Comparative Inference Model (CIM) from paired signals that combine observed tool behavior, scalable supervision from a Bayesian tool-graph simulator, and semantic judgments from LLM-based comparison. The resulting CIM learns to estimate how likely a possible next tool invocation is to support final task success under the current context. Across three tool-use benchmarks and multiple backbone LLMs, CITA consistently improves Tool F1 and task success. Additional analysis shows that CIM learns accurate step-level value estimates for comparative tool choices.

cs.AI↗

arXiv2026

Bayesian power spectrum estimation with modelling of systematic effects in delay-fringe rate space

Observing the Epoch of Reionisation using 21cm radio interferometry has proven to be a challenging task. Extraction of the extremely faint redshifted signal is complicated by the presence of bright foregrounds, radio frequency interference (RFI), and systematic artefacts. We discuss the challenge of accounting for systematic effects, particularly cable reflections, that appear in the visibility data obtained from 21cm interferometers. Cable reflections cause attenuated copies of the foreground signal to appear outside the 'foreground wedge' region in which foreground contamination is supposed to be localised. We build on the hydra-pspec Gibbs sampler to implement a model of the systematics as a multiplicative effect in delay-fringe rate space. We include this model in the inference of the joint posterior distribution, in addition to the 21cm signal, its power spectrum, and foregrounds. This allows the systematics contribution to be marginalised, rather than filtering it out and causing additional signal loss. We demonstrate the method on simulated visibility data for a single baseline, showing that the 21cm delay power spectrum can be recovered well regardless of the location of the systematics in delay-fringe rate space. Our implementation is suitable for modelling other multiplicative factors on the visibilities, e.g. residual gain errors.

astro-ph.CO↗

arXiv2026

Does the VGGT Family Need All Its Layers?

Which layers of a feed-forward geometry model are needed to preserve both camera poses and dense 3D structure? We study layer redundancy in VGGT, $π^3$, and VGGT-$Ω$: 3,018 pruned configurations, scored on seven camera-pose and dense-geometry metrics across four indoor and outdoor datasets. Four findings follow: (i) Removable layers cluster in two redundancy regions: a dominant early region and a narrower late one, while deletions spanning the intervening layers are consistently more disruptive. This recurring pattern holds across models, datasets, and metrics, and contrasts with the middle-to-late redundancy commonly reported in the literature. (ii) Within these regions, we observe that the joint degradation from deleting two intervals is approximately the sum of their individual degradations, reducing the number of model evaluations for pruning search from $O(L^4)$ to $O(L^2)$, where $L$ is the aggregator depth. (iii) We find that CKA provides a cheaper representation-based proxy for interval degradation, offering a practical trade-off between pruning quality and calibration cost. (iv) Closed-form linear calibration recovers accuracy after pruning without end-to-end retraining. A least-squares analysis shows that using a shared map for special and patch tokens generally incurs excess reconstruction loss, motivating token-aware recovery. Recovery maps fitted on just 100 calibration scenes generalize to held-out scenes and unseen datasets. The resulting models reduce aggregator parameters by up to 44% while maintaining accuracy comparable to their intact counterparts. Code and experimental results will be available at our project page: https://xian-bei.github.io/vggt-family-layer-redundancy/

cs.CV↗

arXiv2026

Train Ahead, Distill Back: Bootstrapping On-Policy Self-Distillation for Large Language Models

On-policy self-distillation (OPSD) improves large language models by letting a self-teacher with privileged information provide dense token-level supervision on the model's own trajectories. Yet existing methods typically construct the self-teacher from the current, initial, or slowly averaged policy state, leaving the quality of supervision constrained by the teacher's ability to exploit privileged information. We ask whether the model's own optimization progress can instead be recycled into a stronger self-teacher. In this paper, we introduce Bootstrapped On-Policy Self-Distillation (B-OPSD), which temporarily trains the policy ahead to obtain a future teacher, restores the student to the original policy state, and then uses the future teacher to supervise the restarted student. The future teacher improves supervision in two complementary ways, it can generate more reliable privileged trajectories and, conditioned on them, provide more informative token-level targets along the restarted student's on-policy trajectories. Experiments on mathematical reasoning with Qwen3-4B and Qwen3-8B show consistent improvements over standard OPSD in both settings, including gains from 27.50 to 41.30 and from 48.80 to 64.44 in the rollout-privileged setting. Our findings point to a broader principle for self-improving models that future learning progress can be distilled backward, preserving acquired knowledge while bootstrapping beyond the optimization state that produced it.

cs.AI↗

arXiv2026

From Judgment Quality to Downstream Utility: Rethinking LLM-as-a-Judge for Open-Ended Tasks

LLM-as-a-Judge is increasingly used to evaluate policy responses on open-ended tasks that lack ground-truth answers. Existing work often directly converts the resulting judgments into reward signals for policy training, paying limited attention to intrinsic judgment quality and largely restricting the use of Judges to training-time supervision. We systematically investigate judgment quality and downstream utility by examining both how judgments are elicited and how they are used. For judgment elicitation, we vary the Judge protocol along three dimensions: verdict granularity, critique usage, and evaluation batching. For judgment usage, beyond policy training, we extend Judge to test-time inference through Best-of-N selection, Judge-guided revision, and beam search. We find that, (i) Surprisingly, judgment quality and downstream utility do not always align. (ii) Judge protocol design substantially affects both intrinsic judgment quality and downstream utility. (iii) Judge guidance effectively converts test-time compute into performance gains, with benefits varying across inference strategies. Our results call for a multifaceted evaluation of LLM Judges on open-ended tasks, encompassing intrinsic judgment quality, and downstream utility.

cs.AI↗

arXiv2026

Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence

Latent visual reasoning (LVR) enables multimodal large language models (MLLMs) to perform intermediate computation in continuous latent tokens rather than expressing every reasoning step in words. However, unlike textual CoT, latent reasoning is not directly observable, making it difficult to supervise what latent tokens learn. In this work, we first conduct a thorough analysis of latent-token behavior and identify a latent evidence-credit gap: latent tokens respond only weakly to image perturbations that alter the correct answer. We hypothesize that this issue stems from the lack of explicit supervision during GRPO training. These findings suggest that a final-answer reward provides too little guidance on what visual evidence to preserve or how credit should be assigned across latent tokens. To bridge this gap, we propose ReaLVR, which brings visual-evidence supervision to the model's own free-running latent trajectories. ReaLVR contrasts correct and model-generated wrong answers to determine where stronger supervision is needed, and relevant and mismatched visual evidence to specify what to preserve. Across three model families, ReaLVR consistently outperforms evaluated LVR baselines, achieving the highest five-task average of 63.7% on Qwen2.5-VL-7B. Crucially, we are the first to scale visual reasoning in latent space, showing that our framework continues to deliver robust improvements at frontier model scales up to 235B. Further analyses show more question-sensitive latent-token positions, stronger alignment with relevant visual regions, and greater fixed-context dependence on the most attended latent tokens.

cs.CV↗

arXiv2026

Simulation-Efficient Analog Circuit Yield Optimization via Monte Carlo Zeroth-Order Gradient Estimation

Yield optimization under process variation is expensive because each candidate design must be evaluated across many Monte Carlo SPICE samples. The resulting finite-sample yield is also piecewise constant in the design parameters, providing little local information for optimization. We introduce zeroth-order Monte Carlo stochastic gradient descent (ZO-MC-SGD), a black-box method that converts continuous specification margins into stochastic descent directions. Each update evaluates opposite design perturbations under shared process samples, allowing a small simulation batch to estimate a local direction without differentiating SPICE or fitting a global surrogate model. A Spearman rank-correlation test checks that the margin-based loss orders designs consistently with empirical yield. We prove that the estimator is unbiased for a Gaussian-smoothed surrogate and derive variance and sample-complexity bounds with no explicit dependence on process dimension. Across five analog circuit benchmarks with up to 30 design variables and 42 process variables, ZO-MC-SGD reaches a mean yield of 0.95 on four circuits within 50--200 simulations and the empirical yield ceiling on the fifth. Relative to the best of five black-box and learning-based baselines, it reduces the required simulation budget by up to a factor of eight.

cs.CE↗

arXiv2026

What Do They Fix? LLM-Aided Categorization of Security Patches for Critical Memory Bugs

Open-source software projects are foundational to modern software ecosystems, with the Linux kernel standing out as a critical exemplar due to its ubiquity and complexity. Although security patches are continuously integrated into the Linux mainline kernel, downstream maintainers often delay their adoption, creating windows of vulnerability. A key reason for this lag is the difficulty in identifying security-critical patches, particularly those addressing exploitable vulnerabilities such as out-of-bounds (OOB) accesses and use-after-free (UAF) bugs. This challenge is exacerbated by intentionally silent bug fixes, incomplete or missing CVE assignments, delays in CVE issuance, and recent changes to the CVE assignment criteria for the Linux kernel. While fine-grained patch classification approaches exist, they exhibit limitations in both coverage and accuracy. In this work, we identify previously unexplored opportunities to significantly improve fine-grained patch classification. Specifically, by leveraging cues from commit titles/messages and diffs alongside appropriate code context, we develop DUALLM, a dual-method pipeline that integrates two approaches based on a Large Language Model (LLM) and a fine-tuned small language model. DUALLM achieves 87.4% accuracy and an F1-score of 0.875, significantly outperforming prior solutions. Notably, DUALLM successfully identified 111 of 5,140 recent Linux kernel patches as addressing OOB or UAF vulnerabilities, with 90 true positives confirmed by manual verification (many do not have clear indications in patch descriptions). Moreover, we constructed proof-of-concepts for two identified bugs (one UAF and one OOB), including one developed to conduct a previously unknown control-flow hijack as further evidence of the correctness of the classification.

cs.CR↗

arXiv2026

Adversarial Closed-Loop Curriculum for Evolving Role-Playing Agents

Role-playing agents based on large language models have been widely applied in areas such as personalized assistance and social simulation. Recent RL methods typically train on a fixed scenario pool collected before learning begins. This creates a distributional bottleneck: as the agent improves, the scenarios where it performs poorly also change, while the training distribution remains static. Therefore, we propose AdvRole, an adversarial context rewriting framework that turns role-playing RL into a closed-loop curriculum. AdvRole alternates between an Actor that learns to role-play and a Rewriter that edits character profiles and dialogue contexts into actor-specific hard scenarios. The Rewriter is trained with a performance-gap reward, which favors rewrites that reduce the current Actor's score relative to the original scenario. As a result, the scenario pool evolves with the Actor and continuously targets under-mastered regions of the character-context space. Experiments on three role-playing benchmarks covering English and Chinese, as well as a new multilingual benchmark we release, show that AdvRole consistently outperforms baselines.

cs.AI↗

arXiv2026

Improving depth-resolution, in-plane contrast, and reducing non-uniformity artifacts for wide-angle DBT

Purpose: This work aims to develop an image reconstruction algorithm for wide-angle digital breast tomosynthesis (DBT) that has improved depth resolution and in-plane contrast while reducing non-uniformity artifacts. Approach: The image reconstruction algorithm is an extension of our prior work on sparsity-regularized iterative image reconstruction. The algorithm is performed in two stages as explained in a prior work. The first stage consists of a low-resolution reconstruction that exploits sparsity for quantitative accuracy. In this work, this first stage is augmented with a formulation that includes the estimation of a "background" image, which absorbs low-frequency artifacts that cause image non-uniformity. Results: The new algorithm is demonstrated on three patient cases for which the data are acquired on a wide-angle DBT system. Conclusion: The results on the shown cases indicate that the algorithm design goals have been met, but additional empirical results and task-based assessment are needed to strengthen this conclusion.

math.OC↗

arXiv2026

Projective representation theory of spin space groups

Spin space groups provide the natural symmetry framework for magnetic crystals with weak spin-orbit coupling, but extracting their physical consequences requires a general theory of irreducible representations. The central difficulty is that spin space groups are represented projectively: their factor systems can render lattice translations noncommuting, induce nonsymmorphic actions in momentum space, and modify the projective structure of little cogroups. These effects lie beyond conventional space-group and double-group representation theory. Here, using Mackey's theory of group extensions, we develop a unified constructive framework for all collinear, coplanar, and noncoplanar spin space groups, including antiunitary symmetries and general factor systems. A central technical result is a canonical decomposition of the relevant factor system $ν$ into a translational factor $σ$, a mixed factor $γ$ coupling translations to point-group operations, and a point-group factor $α$. This decomposition makes transparent how ordinary representation theory is modified: $σ$ determines the projective translation algebra and the appropriate Brillouin zone, $γ$ controls the momentum-space group action and can make it nonsymmorphic, and $α$ contributes to the factor systems of little cogroups. On this basis, we construct all projective irreducible corepresentations by induction over momentum-space orbits. The framework identifies which nonsymmorphic momentum-space symmetries can be realized by spin space groups and reveals Brillouin spaces that are compact flat manifolds rather than tori, symmetry-enforced Zak phases, reconstructed high-symmetry momenta and band degeneracies, and new types of quasiparticles. Our results establish the representation-theoretic foundation for systematic studies of weak-spin-orbit-coupled magnetic materials.

cond-mat.mes-hall↗

arXiv2026

Pattern-Aware Virtual Network Embedding Optimization for Cloud Data Centers

The network virtualization (NV) technology has enabled the sharing of multiple resources among virtual networks (VNs) in cloud data centers. One of the key challenges is to allocate resources in real-time for virtual network request (VNR), which is known as online virtual network embedding (VNE). However, the existing online VNE methods do not exploit the multi-dimensional complementary relationship among diverse VNRs, resulting in the fragmentation and waste of substrate resources. In this paper, we propose the pattern matching based online VNE approach by constructing appropriate matching rules among observed patterns to maximize resources utilization. We devise the clustering based VNRs quantization method and conduct rigorous study on the pattern combination filtering problem. Then, we utilize the column generation to solve it and construct the pattern matching rules. Based on the rules, we propose an online pattern matching VNE algorithm with linear worst-case complexity. Evaluation on a 106-server testbed using Alibaba production cluster trace dataset shows that our algorithm achieves close-to-offline performance and more accepted workloads that outperforms traditional designs by 25%-30%.

cs.NI↗

arXiv2026

Bench2Dex: Benchmarking Visuo-Tactile Bimanual Dexterous Manipulation Across Dexterous Hands

Tactile sensing provides contact information that can be difficult to infer from vision alone, but tactile hardware for dexterous hands has not converged to a common design. Dexterous hands differ in finger structure, contact surfaces, and sensor layouts, while simulated tactile signals still differ from measurements produced by physical sensors. These factors make it difficult to study visuo-tactile manipulation across diverse dexterous hands within a consistent experimental setting. We present Bench2Dex, a simulation benchmark for visuo-tactile bimanual manipulation across 12 dexterous hands. We adapt existing robot models with a shared simulated tactile interface that converts local contact geometry into image-like tactile observations. The interface provides a consistent observation format across different hand morphologies without attempting to reproduce the output of a specific physical tactile sensor. Bench2Dex includes 26 bimanual manipulation tasks that involve tool use, articulated-object interaction, and multi-stage manipulation, together with about 1.3K human-teleoperated demonstrations. The benchmark provides synchronized visual, tactile, proprioceptive, action, and object-state observations, together with executable task metrics. For robustness, we group seven perturbation types into invariance axis, where the correct action does not change, and equivariance axis, where the correct action changes together with the perturbation. We evaluate ACT, Diffusion Policy, pi0.5, and GR00T N1.5 on Bench2Dex and report their performance and failure modes. Bench2Dex is meant as a platform for studying visuo-tactile learning across dexterous hands. It does not assume that simulated tactile observations can replace real tactile sensing; it offers a shared setting for algorithm development while tactile hardware and simulation models are still evolving.

cs.RO↗

arXiv2026

Basis Rigidity of the AES S-box and Generic Rigidity of Inversion under Affine Transformations

The AES S-box is constructed from finite field inversion followed by a fixed affine transformation. Since inversion possesses intrinsic Frobenius symmetries among its coordinate realizations, we study how these basis symmetries are altered by outer affine transformations. We first develop a deterministic rigidity criterion for transformed inversion and apply it to the AES S-box. This shows that the linear part of the AES S-box affine transformation alone makes the transformed inversion map basis rigid. We then investigate the corresponding generic problem when the outer invertible linear transformation varies. The existence of a nontrivial linear stabilizer is reduced to a conjugacy problem for semilinear candidates arising from two sided linear equivalences of inversion, which we characterize in terms of relative norms and Frobenius orbits. We also determine the dimensions of the associated centralizer algebras exactly. These structural results imply that, for a uniformly chosen outer linear transformation, the probability that the linear stabilizer is nontrivial is bounded by $2^{-Ω(n^2)}$, with sharper finite dimensional bounds obtained from the exact conjugacy condition. Computational experiments independently verify the AES rigidity result, the conjugacy and centralizer formulas, and the finite dimensional estimates in small dimensions.

cs.CR↗

arXiv2026

Round-the-Clock Sub-Ambient Cooling via Dynamic Sky View Factor Steering

Passive daytime radiative cooling holds significant potential to address sustainability challenges such as the energy-water nexus. However, common static horizontal configurations cannot respond to dynamic changes of environmental conditions such as solar position and cloud coverage, thereby forfeiting the opportunity of optimizing the cooling performance in the 24-h day-night cycle. Here, inspired from the heliotropism of sunflowers, we develop a dynamic sky view factor steering (DSVFS) system. By dynamically steering the emitter to an optimal angle, this system not only relaxes the stringent dual requirement of ultra-high solar reflectivity and infrared emissivity of the radiative emitter, but also maximizes its cooling power round-the clock. Using this DSVFS system, we experimentally demonstrate sub-ambient cooling during a typical hot noon even with a near-blackbody emitter; with a selective emitter, we demonstrate an increase of cooling power by 135% as compared to its static counterpart. Case studies of multiple cities across the world indicate a maximum annual electricity saving of up to 200 kWh/m^2.

physics.app-ph↗

arXiv2026

Directional memory of early spectral selection during spinodal decomposition in finite systems

Cahn-Hilliard phase separation turns small composition fluctuations into coarsening domains. In Fourier space, the pattern forms a ring whose angular intensity favors an axis. We ask whether the axis favored early in an evolution remains related to the axis favored later in the same evolution. We analyze two-dimensional simulations with isotropic initial fluctuations and no imposed direction. We introduce $Q_2$, a normalized angular average over the Fourier spectrum, as a measure of the strength and direction of spectral anisotropy. Linear growth amplifies different wave numbers at different rates, selecting an early axis from the angular imbalances in the initial spectrum. This axis remains correlated with the late axis during coarsening. At a fixed time, shallow quenches show stronger early-late alignment. This advantage disappears when the quenches are compared at similar stages of domain growth. Adding the early $Q_2$ to a prediction based on quench depth, mean composition, and box size reduces the mean absolute error for late $Q_2$ by $3.53\%$. This reduction disappears when only the magnitude of the early $Q_2$ is retained, or when its direction is replaced by the direction from another evolution. The early Fourier intensity therefore carries information about the later state of the same evolving field.

cond-mat.stat-mech↗

arXiv2026

Representation Redundancy and Structural Complexity in Finite-Field Inversion

The representation chosen for a mathematical operation can affect both its algebraic form and its empirical learning difficulty. We study this phenomenon for inversion over $\mathbb F_{2^n}$, with field elements expressed in varying ordered $\mathbb F_2$-bases. We prove that two ordered bases induce the same coordinate inversion map if and only if they belong to the same Galois orbit. Since every orbit has size $n$, the correspondence between ordered bases and distinct inversion maps is exactly $n$-to-one. We then analyze three Boolean formulations of inversion. The reference formulation has algebraic degree $n-1$ and joint ANF leap $1$, the mixed representation formulation has degree $2(n-1)$ and joint ANF leap $2$, and the complete raw formulation has degree at most $3(n-1)$ and joint ANF leap at least $n$. Exhaustive computations agree with the theoretical results and bounds in the cases considered. Controlled experiments with multilayer perceptrons show the same ordering in learning difficulty, while Galois orbit redundancy provides only a limited generalization benefit under the tested conditions. These results show that exact redundancy among representations can coexist with changes in Boolean structure and learning behavior when the representation is exposed as part of the input.

cs.LG↗

arXiv2026

GRZO: Group-Relative Zeroth-Order Optimization for Large Language Model Fine-Tuning

Zeroth-order (ZO) optimization is a memory-efficient alternative to backpropagation for fine-tuning large language models, but its deployment is limited by the high variance of gradient estimation. We propose GRZO, a Group-Relative Zeroth-Order optimizer that draws one pseudo-independent perturbation per mini-batch example and aggregates the per-example losses through group-relative normalization, raising the effective gradient-direction count from one to the batch size at no additional forward cost while preserving inference-level memory. We prove that GRZO is directionally unbiased with variance shrinking proportionally to the batch size, yielding a tighter nonconvex convergence bound than MeZO. Across RoBERTa-large, Llama3-8B, and OPT-13B over multiple tasks, GRZO improves average accuracy on Llama3-8B by $+3.0$ over MeZO at $23\%$ lower peak GPU memory; as a drop-in replacement for the MeZO core, it lifts sparse, low-rank, and quantized ZO variants by $+6.0$ on average.

cs.LG↗