arXiv subjects

Wei Wang

Publications and source records attributed to Wei Wang.

At least 55 records · Page 3Linked to original sources

arXiv2026

UAV Rotatable Antenna under Rotation Delay

Rotatable antennas (RAs) can improve wireless link performance by adjusting their boresight directions. However, existing RA studies generally assume instantaneous antenna reconfiguration, whereas mechanical rotation incurs non-negligible delay in practical deployment. To address this issue, we develop a rotation-delay-aware communication model for finite communication slots, where mechanical rotation directly reduces the time available for data transmission. This model reveals a fundamental tradeoff between the communication gain achieved through RA reorientation and the associated transmission-time loss. Accordingly, we formulate a rotation-delay-aware joint design of RA orientations and transmit beamforming to maximize the minimum effective transmission rate. For the single-user system, under a sufficient condition for strict concavity, the global optimum falls into one of three regimes: no rotation, partial rotation, and full rotation. A closed-form approximation is derived using a second-order expansion. For the multi-user system, we separate rotation-budget selection from fixed-budget max-min signal-to-interference-plus-noise ratio (SINR) optimization. An analytic-gradient zero-forcing (AG-ZF) algorithm is then developed for low-complexity joint RA orientation and rotation-budget optimization. The results show that explicitly accounting for mechanical rotation delay is essential for finite-slot RA-enabled UAV communications, and that the proposed designs effectively balance reconfiguration gain and transmission-time loss while achieving favorable effective-rate performance with low computational complexity.

cs.IT↗

arXiv2026

STRIDE: Evaluating Spatiotemporal Reasoning in Driving Edge Cases

In autonomous driving, perceiving surrounding objects is not enough: failing to understand how they move and interact over time can lead to dangerous decisions. Evaluating vision-language models (VLMs) therefore requires testing whether they connect past spatial observations to future consequences. Yet existing benchmarks often assess edge cases through basic perception questions or test temporal understanding in simple scenarios, leaving complex spatiotemporal reasoning underexplored. We introduce STRIDE (SpatioTemporal Reasoning In Driving Edge Cases), an object-centric benchmark for long-tail, safety-critical driving scenarios. STRIDE covers six task families: spatial perception, spatial understanding, temporal memory, temporal extrapolation, trajectory prediction, and scene-context awareness. It pairs multiple-choice questions probing individual spatiotemporal judgments with related free-form visual question answering (VQA) requiring detailed scene explanations and justifications of driving decisions. We prioritize scenes with unusual object behaviors and complex traffic interactions, generate reference answers from structured dataset annotations through automated procedures, and use limited language-model assistance and human verification to ensure annotation quality. Across 12 proprietary, open-weight, and driving-specialized models, even the best performer, GPT-6-Astra, achieves only 43.4% average multiple-choice accuracy, versus an approximately 20% random-choice baseline. Further analysis reveals that strong performance in describing driving scenes does not necessarily translate to accurate safety-critical judgments, and that models struggle more with reconstructing object motion than recalling past positions. STRIDE thus exposes spatiotemporal driving limitations that aggregate VQA scores can conceal.

cs.CV↗

arXiv2026

Sharp asymptotics for the Lovász number of random circulant graphs

We study the LOVÁSZ number of dense random circulant graphs generated by independent Bernoulli bits indexed by cyclic distances. Although only $O(n)$ independent random inputs determine $O(n^2)$ globally dependent edge indicators, we prove sharp expectation asymptotics along all odd orders, including composite orders, with leading constant one. This resolves a conjecture posed for random circulant graphs in 2025. The proof combines an exact Fourier reduction and complementary duality with a positive interior smoothing that reveals a nonlinear random fixed-point equation. To analyze this equation under the strong dependence of the distance-bit model, we construct canonical tree fields, establish growing-degree Gaussian approximations through polynomial comparison and Wick/Hermite expansions, and transfer the resulting approximation back to the original sign model. A deterministic residual estimate together with an explicit exceptional-event integration yields the sharp expectation asymptotics.

math.PR↗

arXiv2026

Aspire: Practical Batch Queries on Dynamic Encrypted Database with Full Leakage Suppression

Leakage-abuse attacks against encrypted database systems exploit semantic leakage and pose a severe threat to the confidentiality of plaintext data and queries. To address this problem, we propose a zero-leakage encrypted database that rules out leakage-abuse attacks, which fundamentally rely on semantic leakage. We first distinguish between two kinds of zero-leakage security. Ideal zero-leakage forbids the disclosure of any information beyond public parameters, whereas restricted zero-leakage additionally permits the leakage of metadata (such as database size and query timing). We show that restricted zero-leakage is sufficient to thwart leakage-abuse attacks and is achievable in practice. We then formally define adaptive restricted zero-leakage security and present Aspire, an encrypted database that satisfies this notion. Aspire integrates probabilistic batch code and minimal perfect hashing to support efficient batch keyword queries and dynamic updates while revealing only non-semantic metadata. Experimental results demonstrate the practicality and efficiency of Aspire, showing that Aspire can achieve an amortized query latency of less than 1.38 ms (7.53 s for a batch of 5461 keyword queries), with an amortized communication cost of approximately 2.95 KB (16.1 MB in total), on a $2^{22}$ key-value pairs database.

cs.CR↗

arXiv2026

FuTransformer: FP8 prefill and INT4 decode for Llama-style inference on Hopper

Transformer prefill reuses weights across tokens and is dominated by matrix multiplication and activation data movement; single-token decode has limited reuse and is dominated by weight and KV-cache reads. We present FUTRANSFORMER, a CUDA implementation of a Llama-style Transformer block for NVIDIA Hopper GPUs. It applies CODA's GEMM-plus-epilogue approach to inference, adding direct KV-cache placement, folding the fixed RMSNorm gain into weights, writing FP8 output directly with delayed scaling, and supporting prefill without padding. Decode uses INT4 projection weights and an INT4 KV-cache copy. Weights and keys follow Tensor Core fragment layouts for direct dequantization into MMA operands without full intermediate tiles. A block-diagonal Hadamard transform reduces cache quantization error from outlier channels. On one H100, prefill is up to 1.48x faster than compiled PyTorch FP8 and reaches 1.29 PFLOP/s. For gate/up projection with SwiGLU at the 8B shape, our BF16 path is 1.13x faster than CODA. Our FP8 path is 2.02-2.18x faster than the evaluated BF16 CODA path. Across 8B and 70B shapes at batches 1-32, INT4 decode is 1.3-2.4x faster than the faster of PyTorch's BF16 and INT4 blocks at each point: 2.1-3.3x faster than compiled, CUDA-graphed BF16 and 1.3-6.3x faster than the same block using PyTorch's INT4 projection kernel, which targets a single query row and loses to BF16 at batch 32. It is also 31-51% faster than our FP8 path. Complete Llama-3.1-8B decode takes 2.8 ms per step at batch one, versus 8.1 ms for the compiled transformers implementation. Decoding every WikiText-2 token through the INT4 weight path raises perplexity from 6.24 to 7.23 with round-to-nearest weights or 6.99 with GPTQ in the same format; the rotated INT4 cache adds 0.20. The kernels match a quantized reference computation within 0.001 perplexity.

cs.AI↗

arXiv2026

Rethinking Learning from Label Proportions via Moment Matching

Learning from label proportions (LLP) is a weakly supervised setting where training data are grouped into bags with only bag-level label proportions observed, and the goal is to learn a classifier that predicts labels for individual instances. Recently, a large body of methods has emerged; however, most of them rely on the assumption that instance-label pairs are sampled i.i.d. and randomly assembled into bags, an assumption often violated in real-world scenarios. Our experimental results indicate that these methods perform suboptimally under non-random settings. In this work, we adopt a more realistic assumption: bags are sampled i.i.d., and instance labels are conditionally independent given the instances. Under this setting, we show that the label counts follow a Poisson multinomial distribution. Motivated by this observation, we propose LLP via moment matching (LLP-MM), a simple yet effective approach which leverages multi-order factorial moments as training objectives, encouraging the classifier predictions to match these moments and thereby more fully exploit the underlying statistical structure of the data. Extensive experiments on benchmark datasets under various bag construction strategies demonstrate the effectiveness of our approach while maintaining high computational efficiency. Our code is available at https://github.com/TianhaoMa5/LLP-MM.

cs.LG↗

arXiv2026

Mem4Act: Learning Effective Latent Memory for Long-Horizon Robotic Manipulation

Long-horizon robotic manipulation fundamentally requires memory: agents must retain task-relevant information from past observations and interactions and use it to guide future actions. Existing approaches typically preserve raw historical observations, compressed visual tokens, or language summaries, but these representations either expose the action policy to increasingly long histories or risk discarding task-relevant temporal, semantic, and perceptual information. We introduce Mem4Act, a framework for learning effective latent memory as a fixed-size representation explicitly optimized for long-horizon robotic control. Mem4Act addresses three coupled challenges: how to efficiently cover long histories, what information memory should retain, and how to ensure the policy actually relies on memory. To this end, it organizes history via multi-granularity temporal aggregation, preserves complementary semantic and perceptual information through question-answering supervision and key-frame DINO feature reconstruction, and employs memory-first two-stage training to encourage greater use of memory by the action policy. The resulting memory captures both high-level task state and fine-grained perceptual evidence while providing a bounded interface to the action policy. Across RoboMME, RMBench, and four real-world manipulation tasks, Mem4Act achieves state-of-the-art performance on memory-dependent control, with success rates of 49.88%, 61.0%, and 81.5%, respectively. Extensive ablations further demonstrate the importance of temporal aggregation, complementary memory supervision, and memory-first training.

cs.RO↗

arXiv2026

Lean Graph Search for Formal Theorem Proving

Retrieving relevant lemmas is crucial for formal theorem proving, enabling provers to build machine-checkable proofs from established mathematical results. Recent retrieval systems match library declarations to a target theorem or its intermediate proof steps, but make limited use of multi-hop dependencies among declarations. These dependencies can reveal supporting lemmas that are not necessarily lexically similar to the target theorem but recur in proofs of related problems. To this end, we propose LeanGraphSearch, a graph-based retrieval method for Lean 4 that constructs a graph from Mathlib's forward dependencies and traverses it to uncover supporting lemmas through multi-hop dependency chains. Specifically, LeanGraphSearch links each declaration to the premises it depends on, uses a graph-based ranking algorithm to propagate relevance from semantically matched declarations to their dependencies, and reranks the expanded candidates against the original query. Across two mathematical retrieval benchmarks, LeanGraphSearch improves recall by up to 12% over the strongest baseline, LeanSearch v2, on MathlibQR and increases coverage of complete annotated proof routes on MathlibMPR, covering up to 3.7% of extra groups. More importantly, we evaluate whether these retrieval gains translate into improved end-to-end theorem proving by integrating LeanGraphSearch into an iterative proving loop guided by Lean feedback. Across three theorem proving benchmarks, LeanGraphSearch improves pass rate at {8, 16, 32} reflection turns by as much as 40% compared to vanilla prover and at most 10% compared to LeanSearch v2, showing higher accuracy-cost efficiency. These findings highlight the value of dependency graph search for theorem proving and motivate further research on structure-aware search for proof discovery.

cs.LG↗

arXiv2026

DriveWorld3D Evidence Constrained 3D World Modeling and Generative Simulation for Real-World Autonomous Driving

Three dimensional modeling for autonomous driving must continuously estimate road geometry, obstacle occupancy, traffic participant behavior, and observational uncertainty under limited computational resources. Existing reconstruction methods can synthesize realistic novel views, but visual fidelity does not sufficiently characterize the geometric reliability required for driving. Generative world models can expand scenarios and predict future observations, yet may introduce content unsupported by measurements. We propose DriveWorld3D, a framework combining temporal video learning, measurement constraints from multiple sensors, and offline generative simulation. Its online component maintains a unified world state comprising road surfaces, volumetric occupancy, dynamic entities, and road relations, while separately representing physical occupancy, observability, and evidence provenance. Action conditioned future prediction and risk sensitive planning connect this representation to driving decisions. Its offline component reconstructs recorded scenes using neural radiance fields or 3D Gaussian Splatting, then expands long tail conditions using generative models while restricting the use of generated content as geometric ground truth. We specify data partitioning, sensor ablations, fault injection, paired closed loop testing, and real to simulation consistency assessment. The central hypothesis is that explicit evidence management can reduce dangerous false free space predictions and occlusion risk within a matched computational budget while retaining acceptable driving efficiency. All performance claims remain subject to experimental validation. Keywords: autonomous driving; 3D occupancy; sensor fusion; uncertainty; generative world models; NeRF; 3D Gaussian Splatting; closed loop simulation

cs.RO↗

arXiv2026

Mothership: A Two Tier Autonomous Refuse Collection System for Pedestrian Campuses

Refuse collection in pedestrian campuses, parks and mixed-use districts remains manual because compaction trucks cannot reach bins located on walkways, courtyards and building service doors. This paper presents Mothership, a two-tier autonomous refuse collection system in which an electric compaction truck (the mothership) carries two low-speed sidewalk robots (RoboCarts) that fetch standard two-wheel bins, deliver them to the truck, return the empty bins to their original positions and re-dock on the truck for charging. The architecture separates responsibilities by scale: the truck performs road driving, compaction, lifting and energy supply, while the carts perform only sidewalk transport and bin grasping. This split removes the lift-and-tip mechanism from the sidewalk vehicle, keeping it light and low, and relaxes cart-to-truck docking tolerance to the +/-100 mm capture window of a standard automated lifter. We describe the system architecture, the bin-handling and docking interfaces, and a mission-planning formulation that decomposes a route into truck stop points and cart tasks. A discrete-event simulation over synthetic campus layouts with 2-16 bins/ha quantifies throughput as a function of bin density, cart count and stop-point radius. A two-cart configuration serves 20-29 bins/h. Throughput is governed by an interaction between cart count and stop-point radius: a single cart performs best with many close stops, whereas three or four carts need wide stops (80 m) to stay busy, reaching 35-49 bins/h at moderate-to-high densities. The best stop radius therefore has to be chosen together with the fleet size, not fixed in advance.

cs.RO↗

arXiv2026

REVEAL: Reverse Engineering Agentic Latents from Frontier-Model Traces

Frontier LLM agents perform complex interactive tasks, but their inference costs motivate transferring their experience to smaller open-source agents. Distillation and latent communication typically require target-model training or access to source-model internals. We study how textual frontier-model trajectories can provide reusable local controls for a frozen agent. Inspired by protocol reverse engineering, we introduce $\textit{Agentic Latents}$, comprising anchored states and dynamic profiles of feedback--response cycles elicited by replay in the target model. Our training-free framework, $\textbf{REVEAL}$ ($\textbf{REV}$erse $\textbf{E}$ngineering $\textbf{A}$gentic $\textbf{L}$atents), uses these profiles and trajectory outcomes to match and select local events from successful and failed rollouts of the same task and source model. Their response-transition contrasts yield a target-specific steering direction, applied before each response. Within-benchmark evaluations on Terminal-Bench 2.1 and Humanity's Last Exam with tools report higher accuracy than unmodified agents across three targets. Representation analyses and behavioral case studies support reusing frontier-model experience as local activation control.

cs.AI↗

arXiv2026

Risk-aware Road-network Trajectory Synthesis under Local Differential Privacy

Road-network trajectory synthesis under local differential privacy (LDP) enables mobility data analysis without uploading raw trajectories to an untrusted server. Existing locally private trajectory synthesis methods mainly model whole trajectories, point sequences, or grid-level mobility patterns. However, they cannot handle non-uniform privacy risks caused by local road-network and trajectory contexts, such as endpoints, long-stay-associated regions, and low-out-degree branches. They also do not directly protect conditional next-hop choices on road networks, where a user's selected successor may reveal sensitive local movement intentions. To address these challenges, we propose TrajRACE, a locally private synthesis framework for road-network trajectories. TrajRACE uses start, segment-count, and transition reports to represent trajectories. For each transition, it keeps the current road segment as a public aggregation context and perturbs the true successor among legal successors on the client side. It then assigns risk-aware protection strengths to transitions according to local road-network and trajectory contexts. Finally, the server recovers population-level start, segment-count, and conditional transition statistics from heterogeneous privatized reports, and generates synthetic trajectories that follow the public road-network topology. Experiments on three road-network trajectory datasets demonstrate the effectiveness of TrajRACE in improving synthetic-data utility across distributional, query-based, and pattern-level metrics, while providing stronger protection for high-risk local transitions.

cs.CR↗

arXiv2026

When Does Depth Help? Understanding Layerwise Inference in Tabular Foundation Models

Transformer-based tabular foundation models (TFMs) achieve state-of-the-art performance on small to medium-sized tabular prediction tasks while processing diverse inputs through a shared stack of layers. Recent layerwise analyses identify strong predictive performance in early TFM representations and substantial overlap in computation across layers. However, early predictive accessibility alone does not establish whether later layers are dispensable for every input, and aggregate accuracy can obscure which inputs benefit from additional computation. We study all 24 in-context-learning layers of TabPFN-3 across 272 classification datasets, combining representation geometry, prediction trajectories, truncation, and structural interventions, while also examining related representation dynamics in four additional TFMs. We uncover a pronounced expansion-compression trajectory, with stronger late dimensional contraction associated with larger increases in class-prototype separation. A targeted analysis of early-stabilizing samples shows that substantial representational changes persist after predicted classes stabilize, while locally ambiguous queries obtain larger accuracy gains from shallow to late layers than well-separated queries. Repeating a single pretrained block or perturbing layer order reduces predictive performance on average, constraining simple post-hoc substitutions in the frozen model. These findings distinguish predictive accessibility, class stability, and the benefits of continuing computation, supporting an input-dependent assessment of depthwise redundancy in TFMs.

cs.LG↗

arXiv2026

Optimal Design of Shaped Pupil Lyot Coronagraph for China's Large Segmented Telescope

China's next-generation large telescopes, such as the Expandable Aperture Segmented Telescope (EAST) and the Large Optical/Infrared Telescope (LOT), are being planned as flagship facilities for high-angular-resolution astronomy, including the direct imaging and characterization of exoplanetary systems (e.g., exoplanets, circumstellar disks). However, their segmented, centrally obscured apertures present significant challenges for coronagraph design: diffraction from segmented gaps and spiders severely limit contrast, particularly at small inner working angles (IWAs). To address this, we investigate an optimal Shaped Pupil Lyot Coronagraph (SPLC) architecture tailored for these telescopes, which combines a binary shaped pupil apodizer, a focal-plane occulting mask, and a Lyot stop to suppress stellar glare and enhance contrast. We further augment the coronagraph design with a two-deformable-mirror (2-DM) adaptive optics (AO) system to correct atmospheric turbulence and residual aberrations. End-to-end simulations based on site parameters for Lenghu in Qinghai Province and assuming idealized extreme AO (ExAO) correction demonstrate that the optimized 2DM-SPLC designs can achieve raw contrasts of order 10^-6-10^-7 in the V, H, and K bands, providing an upper-bound estimate of the system performance. The simulated performance highlights the advantages of the SPLC in mitigating diffraction from segmented and centrally obscured pupils while maintaining a small IWA and practical throughput, indicating that the 2DM-SPLC is a promising coronagraphic architecture for high-contrast exoplanet imaging on China's future large segmented telescopes.

astro-ph.IM↗

arXiv2026

How Narrative Wrapping Affects LLM Refusal: A Cross-Language Benchmark and Defense

Safety-aligned language models often refuse a harmful request stated directly but answer the same request inside a role-play or narrative wrapper. We measure this vulnerability across languages and registers: attack success on Qwen3-1.7B is already 89.4% in English and 93.0% in modern Chinese, and reaches 95.7% in Classical Chinese. We build GUISE, a benchmark for systematically studying this vulnerability. It includes parallel requests in English, modern Chinese, and Classical Chinese, matched harmful and benign pairs, wrapper types held out for evaluation, and a stricter criterion that counts warn-then-answer responses as attack successes. Representation analysis shows that language and register move harmful-request representations only slightly away from the model's refusal direction, whereas narrative wrappers move them much farther away. We propose AXIS, which combines preference optimisation with a rotation objective that aligns harmful-request representations with the refusal direction and a commitment objective that trains the model to refuse completely rather than produce a warn-then-answer response. Across Qwen3-1.7B, Qwen3-4B and GLM-4-9B, AXIS achieves the highest combined safety and usability score among the compared methods.

cs.AI↗

arXiv2026

On the generalized cospectral mates of chain graphs of rank 4

A graph is called a chain graph if it is bipartite and the neighborhoods of the vertices in each color class are linearly ordered by inclusion. We study the generalized spectral determination of chain graphs of adjacency rank $4$. We first show that generalized cospectral chain graphs have the same number of isolated vertices, reducing the problem to the connected case. For connected rank-$4$ chain graphs, we derive four explicit equations in the cell parameters that are necessary and sufficient for generalized cospectrality. These equations reduce to a quadratic equation in a single parameter. As a consequence, every connected rank-$4$ chain graph has at most one generalized cospectral mate, and the possible mate can be reconstructed explicitly. We further obtain a simple sufficient condition for a rank-$4$ chain graph to be determined by its generalized spectrum (DGS) and, in particular, prove that every double star graph is DGS. Finally, we construct an explicit infinite family of non-double-star rank-$4$ chain graphs that are DGS but not determined by their adjacency spectra, with both the graphs and their complements being connected.

math.CO↗

arXiv2026

Can Agents Forget When Asked? A Benchmark for Memory Unlearning in LLM Agents

Long-term memory has become a core component of large language model (LLM) agents. As agents accumulate personal information over long-term use, users increasingly ask them to delete sensitive or outdated facts. A memory system should honor such requests. It should remove exactly what was requested and keep the rest of its memory intact. However, existing memory benchmarks evaluate only memory writing, retrieval, and organization, so whether agents can forget at the user's explicit request remains under-evaluated. To address this gap, we propose Bemula, the first benchmark dedicated to memory unlearning. It comprises 50 multi-session dialogue instances, 135 forget requests of six types, and 1,742 test questions. Each dialogue has a control version without the requests, so that only systems that have stored the target can earn credit for forgetting. Across the paired versions, five test suites measure storage, removal of the requested information, preservation of unrelated memories, relearning of similar facts, and resistance to extraction attacks. On Bemula, we evaluate six mainstream memory systems under three backbone LLMs, together with the Full Context LLM on two of them. The results are negative. Across all evaluated systems and backbones, no system passes more than 1.5% of forget requests end to end, and most pass none at all. These systems either leave the requested information extractable under simple attacks or delete far more than requested. Memory unlearning therefore remains an open challenge for agent memory design. We release our data and code at https://github.com/Nboxff/bemula-benchmark.

cs.AI↗

arXiv2026

GeoMotionGPT: Geometry-Aligned Motion Understanding with Large Language Models

Discrete motion tokenization has recently enabled Large Language Models (LLMs) to serve as versatile backbones for motion understanding and motion-language reasoning. However, existing pipelines typically decouple motion quantization from semantic embedding learning, linking them solely via token IDs. This approach fails to effectively align the intrinsic geometry of the motion space with the embedding space, thereby hindering the LLM's capacity for nuanced motion reasoning. We argue that alignment is most effective when both modalities share a unified geometric basis. Therefore, instead of forcing the LLM to reconstruct the complex geometry among motion tokens from scratch, we present a novel framework that explicitly enforces orthogonality on both the motion codebook and the LLM embedding space, ensuring that their relational structures naturally mirror each other. Specifically, we employ a decoder-only quantizer with Gumbel-Softmax for differentiable training and balanced codebook usage. To bridge the modalities, we use a sparse projection that maps motion codes into the LLM embedding space while preserving orthogonality. Finally, a two-stage orthonormal regularization schedule enforces soft constraints during tokenizer training and LLM fine-tuning to maintain geometric alignment without hindering semantic adaptation. Extensive experiments show that our framework improves the aggregated Average by 22.4% over the strongest baseline on HumanML3D and by 14.4% on KIT-ML, while ablations confirm the effectiveness of the tokenizer, projection, and regularization designs.

cs.CV↗