Search arXiv⌕ Search

arXiv subjects

Yuhao Zhang

Publications and source records attributed to Yuhao Zhang.

At least 19 recordsLinked to original sources
arXiv2026

Maestro: Agentic Formal Proof Development by Coordination

Lean is increasingly used to verify mathematical proofs and software produced by LLM agents, but writing formal proofs remains hard. Existing prover systems take a divide-and-conquer approach that recursively decomposes hard proof goals into easier lemmas and splits again any lemma that resists proof. However, a lemma can resist proof for different reasons, including insufficient guidance, a false statement, or a flaw in the decomposition that produced it. Existing systems cannot tell these apart as a failed attempt reports little beyond its failure, and responses are confined to the failing node without considering the rest of the proof or revising the plan above it. The system may thus keep decomposing obligations along a flawed proof direction until the budget runs out, causing the whole proof attempt to fail. We introduce Maestro, a Lean prover harness that handles failures through global coordination. Provers report why they failed, not just that they failed, and a coordinator that sees the whole proof chooses an appropriate response, such as guiding a stuck prover, correcting a false lemma, or retracting a flawed subtree and replanning it. The same harness extends to verified code synthesis, where a checked counterexample sends the code back for repair. We evaluate Maestro with four frontier LLMs on seven mathematics and three software verification benchmarks. It proves more targets than each model's native coding agent (Claude Code or Codex) on every benchmark the agent has not already saturated, with the largest gains on research-level mathematics. With Claude Opus 5, it discovers its own proofs of 34 of the 48 theorems in the June 2026 release of arXivLean, against 19 for the native agent. It also outperforms all published prover systems on every benchmark they share, solving 669/672 PutnamBench, 59/60 Lean-IMO-Bench, and 6/6 IMO 2026.

cs.LG↗
arXiv2026

Hilti-Trimble-Oxford Dataset: 360 Visual-Inertial Benchmark with Floor Plan Priors for SLAM and Localization

Automated progress monitoring on construction sites is an active area of research and development. Robot and human-carried mapping systems have been developed to build 3D maps of building and infrastructure projects. While LiDAR-based mapping systems achieve high accuracy, the cost of LiDAR can be prohibitive. Consumer-grade cameras with wide field of view ("360 cameras") combined with embedded inertial measurement units (IMUs) provide a cost-effective alternative. To support change detection and progress monitoring, highly accurate visual simultaneous localization and mapping (SLAM) and floor plan localization systems are required. In this paper, we present a high-quality dataset collected at an active construction site, which captures realistic challenges such as variable lighting conditions, moving workers, fast motions, and repetitive structures. The dataset offers thirty visual-inertial sequences recorded across ten floors over an eight-month period of the construction project. Ground truth trajectories are collected using a high quality LiDAR-inertial SLAM system rigidly attached to the 360 camera. Additionally, we report the results of an open research challenge evaluating the best visual SLAM and localization systems from around the world. The Challenge attracted substantially higher participation in SLAM, with 62 teams compared to 22 in floor plan localization, which suggests that this task receives less attention from researchers and motivates further exploration through proposed dataset and challenge, available at: https://hilti-trimble-challenge.com/dataset-2026.

cs.RO↗
arXiv2026

FDE-Gym: The Last Mile of AI Delivery

Existing agent benchmarks typically evaluate agents on well-specified tasks defined by issue reports, specifications, or task tickets. Real-world customer delivery is different: forward-deployed engineers (FDEs) must clarify ambiguous requirements, work in live systems, and uncover unstated constraints. We introduce FDE-Gym, a benchmark of 95 end-to-end delivery tasks from 20 customer scenarios across 18 sectors, each adapted by industry experts from real-world delivery projects. In simulated customer environments with scripted stakeholders and modifiable customer systems, agents interact with stakeholders and operate on production-like systems to deliver working solutions. FDE-Gym scores a delivery by how much it improves the customer's own success metric, normalized between 0 (unchanged system) and 1 (reference solution), with penalties for stakeholder interactions (messages, trial runs, and waiting time). An online platform lets humans practice FDE skills under the same grading criteria. We evaluate 13 model configurations from seven providers. In the realistic setting, where unstated constraints must be discovered, the best configuration achieves a mean FDE-Gym score of 0.56 and resolves only 14 of the 95 tasks (a score of at least 0.8 with no incident). Discovering unstated constraints is the primary bottleneck: performance drops sharply when models must uncover them without explicit guidance. Yet information discovery alone does not account for the gap: even on the 23 tasks where the best configuration obtains every key fact from stakeholders, it reaches a normalized business uplift of 0.8 or above on only 10. These results reveal a substantial gap between solving well-specified tasks and delivering effective AI systems for real customers.

cs.AI↗
arXiv2026

ContextWeaver: Selective and Dependency-Structured Memory Construction for LLM Agents

Large language model (LLM) agents often struggle in long-context interactions. As the history grows, sliding windows and prompt compression may drop earlier information that later steps rely on. Retrieval-based memory systems surface relevant content but still overlook the causal and logical dependencies between steps. We introduce ContextWeaver, a selective and dependency-structured memory framework that organizes an agent's interaction trace into a graph of reasoning steps and selects the relevant context for future actions. ContextWeaver supports (1) dependency-based construction and traversal that link each step to the earlier steps it relies on; (2) compact dependency summaries of the reasoning paths; and (3) a lightweight validation layer that incorporates execution feedback. On five software-engineering and terminal benchmarks with a fixed budget per task, ContextWeaver improves the resolve rate over a sliding-window baseline on all five, with a significant gain when pooled across benchmarks, while using 13-18% fewer agent steps. It also performs favorably against alternative context-management approaches, with the largest gains on long-horizon repository tasks. On SWE-Bench Pro, ContextWeaver raises the resolve rate from 66.5% to 72.6%, with significant improvements across all three model backbones. The improvement also holds when the sliding window's budget is raised to cover the summary cost.

cs.CL↗
arXiv2026

Verifier-Grounded Agentic RL for Provably Correct Code

Reinforcement learning for code usually rewards passing tests, which a model can do while remaining wrong on the inputs the tests do not cover. A program verifier instead decides whether a program satisfies a formal specification for all inputs. Language models are almost always paired with verifiers at inference time only, and turning a verifier into a reward raises two difficulties. It reports only acceptance or rejection, which gives no gradient when every attempt fails, and a policy can cheat it, most simply by telling it to assume what the proof was supposed to establish. We build an agentic environment and reward that address both. The reward reads the verifier's diagnostics rather than its verdict, scoring a candidate by the weighted fraction of proof obligations it discharges, extended with hidden probes from a reference proof's annotations. An integrity audit outranks the verifier, rejecting any candidate that alters the specification or assumes what it was meant to prove. The model works as an agent against a live verifier over many turns, with a separately trained proof specialist as a frozen collaborator. Across Qwen3 models from 1.7B to 14B and a 35B Qwen3.6 mixture-of-experts model, held-out verification on code generation rises from 66.1% to 93.5% on DafnyBench and from 56.3% to 100% on SWE-Proof. This reward leads every other training signal we compare, including a binary pass-or-fail reward trained under otherwise identical settings, with the widest margin on code generation. When the agent must also deliver the repository patch, graded by each project's withheld tests, the fraction of held-out issues it resolves rises from 59.6% to 72.5%. Evaluated unchanged on three verification suites withheld from training, the checkpoints improve on all of them, including one whose language and verifier are absent from training.

cs.LG↗
arXiv2026

A $1/2$-Approximation for Pandora's Problem with Deadlines

Pandora's problem is a fundamental model of sequential search in which opening each box incurs a cost. We study the variant with deadlines, where each box has a publicly known deadline for both opening and selection. We prove that a simple threshold strategy achieves a $1/2$-approximation to the optimal expected utility, improving the previous guarantee arbitrarily close to $1/4$ due to Amanatidis et al. (2025).

cs.GT↗
arXiv2026

FoundationVIO: Enforcing Metric Scale in Feed-forward Visual-Inertial Odometry

Feed-forward geometric foundation models (GFMs) directly predict dense geometry and camera poses from batches of images. Emerging visual SLAM methods chain together consecutive GFM batches to form dense point cloud maps. A core challenge is how to address the ambiguous geometric scale of GFM predictions: independent predictions are locally consistent up-to-scale but not constrained by any direct physical measurements. Models trained on data with known scale improve the consistency of GFM reconstructions but again do not directly measure it. Consequently, trajectories and 3D reconstructions from GFM-enabled SLAM methods inherit this issue. Motivated by parallels with classical visual-inertial odometry, we introduce FoundationVIO which integrates inertial sensing with GFM predictions to recover metric scale reconstruction using a factor-graph based approach. We demonstrate odometry estimates competitive with classical state-of-the-art VIO methods and accurate dense reconstructions that outperform GFM-enabled SLAM methods. We also add a minimal GFM-based loop-closure mechanism to create a full Visual-Inertial SLAM system capable of creating globally accurate reconstructions and pose estimates. We evaluate our method on three datasets with well known sensor systems and demonstrate its compatibility on ubiquitous camera and IMU sensors. The method runs at approximately real-time on a consumer-grade laptop GPU. Code: https://github.com/ori-drs/FoundationVIO.

cs.RO↗
arXiv2026

SkillScriptBench: Benchmarking Self-Evolution of Executable Agent Skill Packages Beyond Markdown

Executable Agent Skills combine natural-language instructions and scripts into reusable packages for LLM agents, and revising them requires fixing errors without breaking correct behavior. Existing benchmarks do not systematically distinguish documentation repair, script repair, and preservation when evaluating skill self-evolution. We introduce SkillScriptBench, a 350-task benchmark designed to evaluate these capabilities separately. From a survey of over 35,000 GitHub-hosted Skill roots, we select 100 packages and construct 150 repair tasks. Each task pairs a package containing injected script faults with a maintenance request and executable checks of the required behavior. A complementary controlled track contains 200 tasks from 50 packages, each evaluated under the same maintenance request in four states: clean, documentation faults, script faults, and faults in both. Across four LLMs, methods that edit both documentation and scripts can repair script faults but do not consistently outperform Markdown-only revision on documentation repair or preservation. We therefore introduce AST-Guided Skill Revision, which uses abstract syntax trees and calling relationships to link maintenance requirements to relevant code locations. It restricts script edits to these locations and updates the documentation to match the revised scripts. Averaged across models, this revision stage yields absolute gains in repair success of 21.9% for Raw Package and 27.7% for CoEvoSkills on faulty packages. Absolute gains in the proportion of tasks solved in all three runs reach 20.8% and 31.5%, respectively, indicating more consistent repair success across repeated runs.

cs.AI↗
arXiv2026

SDPAD: A Fully Spike-Driven Pipeline for End-to-End Autonomous Driving

End-to-end autonomous driving demands trajectory planners that are both highly accurate and cheap enough for edge deployment. State-of-the-art artificial neural network (ANN) planners meet the accuracy requirement at the cost of heavy dense computation, while spiking neural networks (SNNs)---though promising orders-of-magnitude energy savings through sparse, event-driven arithmetic---still lag far behind in planning accuracy. We present \textbf{SDPAD}, a fully spike-driven end-to-end planning pipeline that closes this gap. SDPAD converts a pre-trained ANN perception stack into integer-spike form via quantized ANN2SNN conversion, lifts multi-view images into the bird's-eye-view (BEV) space with a spike-driven-max (SDM) depth distribution (Spike-3D-Lift), and plans through the Spike-QFormer, a spiking query transformer in which ego, agent, and map queries distilled from the BEV scene are fused by learnable waypoint queries via cross-attention, followed by deformable spike-cross-attention refinement. Every operation is gated by integer spikes and inference is a single feed-forward pass without temporal simulation loops. On the nuScenes open-loop benchmark, SDPAD achieves an average $L_2$ error of 0.40\,m and a collision rate of 0.12\%, on par with strong ANN planners while consuming 69.9\,mJ---less than 2\% of recent ANN baselines. In closed-loop evaluation on the NAVSIM navtest split, SDPAD reaches 86.3 PDMS, surpassing the previous SNN planner SAD by 4.3 points and matching mainstream ANN planners at a fraction of their energy. To our knowledge, SDPAD is the first fully spike-driven planner evaluated in end-to-end autonomous driving, demonstrating that SNNs can rival dense ANNs in complex driving tasks.

cs.RO↗
arXiv2026

Localization in Multi-Panel Massive MIMO With Clock Asynchronism: A Unified Approach

In this work, we propose a unified localization framework, termed UNILocMP, that combines model-based geometry and channel charting (CC) for multi-panel massive multiple-input-multiple-output (MIMO) under clock asynchronism. Owing to the multi-panel architecture, users are classified into multi-line-of-sight (LoS) users, which maintain LoS links with at least two panels, and single/non-LoS users, which maintain a LoS link with only one panel or with none of the panels. For multi-LoS users, a joint position and clock bias estimation is developed; while for single/non-LoS users, an unsupervised CC model is trained with a two-stage data augmentation strategy, where a clock-bias aware dissimilarity metric is introduced. It is numerically validated that the proposed UNILocMP outperforms model-based and CC-based baselines and achieves acceptable performance compared with fully-supervised fingerprinting. Moreover, for a fixed total number of antennas, the multi-panel architecture significantly improves localization accuracy and robustness compared with a single-panel base station (BS) deployment, particularly in the presence of clock asynchronism.

eess.SP↗
arXiv2026

Functionally Equivalent or Not? Graph-Grounded Differential Surrogate Execution for Code Equivalence

Determining whether two programs are functionally equivalent is central to code modernization, patch validation, refactoring, and code-generation evaluation. Yet the usual signals are incomplete: tests cover only finite inputs, textual similarity confuses implementation with behavior, and unconstrained LLM judgments are difficult to audit. Direct execution is often impossible when a program depends on an obsolete, licensed, unavailable, or unsafe environment. We introduce FEAgent, a selective equivalence assessor agent that combines typed program-graph evidence with differential surrogate execution. FEAgent first aligns public interfaces and behaviorally relevant graph anchors, then issues bounded queries over call-flow, control-flow, data-flow, type, import, and effect relations. Next, a branch-aware generator agent proposes discriminating inputs, and two blinded LLM surrogates independently predict source and target observables. Every claim and predicted divergence is recorded in an evidence ledger. A deterministic reconciler then returns EQUIVALENT, INEQUIVALENT, or UNCLEAR rather than forcing a verdict when paths are uncovered or evidence conflicts. We evaluate FEAgent on function-level equivalence and repository-level bug patches, where the existing oracle is a benchmark label or a passing test suite. Every disagreement with that oracle is adjudicated by direct execution, revealing errors in benchmark labels and behavioral divergences missed by unit-test-only scoring. On EquiBench, execution confirms FEAgent's disagreements with published labels on 216 of 1,200 evaluated pairs (18.0%); on SWE-bench Verified, 94 of 331 test-passing agent patches (28.4%) diverge from the reference patch. FEAgent thus serves as an audit layer between testing and formal verification, keeping its evidence reviewable and its uncertainty explicit without claiming a proof of equivalence.

cs.AI↗
arXiv2026

HyperLogic: A Hard, Forward-Authored Chinese Logical Reasoning Benchmark with Execution-Derived Answers

Existing logic benchmarks primarily measure models' ability to answer reasoning questions directly. Scalable benchmarks often generate text from formal structures, which makes answers easy to compute but fixes the formalization before the problem is written. Forward construction preserves the challenge of finding a faithful formalization, yet makes difficulty and answer reliability harder to control. We introduce HyperLogic, a forward-construction pipeline that separates problem authoring from answer generation. A multi-agent workflow hardens undergraduate-authored Chinese seeds without solving them; two agents from different model families independently translate each finished item into executable finite-domain models; their encodings and solver-derived answers undergo layered, agent-assisted adjudication under human-expert oversight. HyperLogic-Base contains 195 items and 922 sub-questions and separates seven frontier models by 33.0 percentage points in strict item accuracy (44.6-77.6%). HyperLogic-Hard contains 100 items with larger, coupled search spaces, on which no model exceeds 16% accuracy in direct answering. We also use Hard to evaluate agents' ability to formalize and solve problems with tools, comparing a code sandbox alone with one that includes our logic modeling library. The sandbox improves every model by 16.7-40.1 points; adding the library helps five models and hurts two. These results highlight the difficulty of faithful formalization even with tool access.

cs.CL↗
arXiv2026

Truthful-in-Expectation MMS Allocations for Chores

We study truthful-in-expectation (TIE) mechanisms for allocating indivisible chores alongside ex-post maximin share (MMS) guarantees. For goods, Bu and Tao (FOCS 2025) established a (1/n)-approximation for TIE mechanisms, and this was substantially improved by Babaioff, Feige, and Manaker Morag (FOCS 2026), who established an Ω(1/\log n) approximation, where n is the number of agents. The corresponding problem for chores has received less attention. The best-known result is due to Aziz, Li, and Wu (MAPR 2024), who gave a TIE mechanism with an O(\sqrt{\log n}) MMS approximation guarantee that holds only in expectation. They also established a 6/5 lower bound for TIE mechanisms for two agents. In this paper, we present the first TIE mechanism for chores that achieves a constant ex-post MMS approximation guarantee. Specifically, our mechanism guarantees an ex-post ratio of 1.97 for any number of agents n, which improves to 4/3 for n=2 and 3/2 for n=3. On the hardness side, we tighten the two-agent lower bound to 4/3, showing that our mechanism is optimal for n=2. More generally, we establish a lower bound of 13/12 on the ex-post MMS approximation ratio achievable by TIE mechanisms for every n\ge 3.

cs.GT↗
arXiv2026

Denoising Multi-Robot Trajectories

Multi-robot trajectory planning is a fundamental problem in multi-robot coordination but remains computationally challenging due to its nonconvex, multimodal, and high-dimensional nature. This work builds upon D4orm, a dynamics-aware diffusion-denoising framework, and develops a family of planning architectures for diverse operational requirements. Unlike conventional numerical optimization methods, D4orm employs sampling-based optimization to generate solution trajectories through massively parallel sampling, leveraging modern computing architectures such as GPUs. Its diffusion-denoising structure iteratively optimizes \textit{deformations} to candidate control trajectories, providing an efficient and versatile paradigm for generating kinodynamically feasible and conflict-free trajectories. Using D4orm as the building block for advanced planners, we present a decoupled planner for improved scalability, an online receding-horizon planner with feedback control, and a distributed planner for resource-constrained settings. Evaluations with differential-drive and holonomic robots in 2D and 3D environments demonstrate that D4orm-based approaches find high-quality solutions faster and more reliably than other sampling-based optimization methods, such as MPPI, as well as a learned diffusion-model-based method. We further demonstrate zero-shot deployment on ten real quadrotors with obstacles, large-scale deconfliction with 100 simulated robots, and fully onboard distributed `lifelong' operation with six ground robots. Overall, these results establish diffusion denoising as a scalable and reliable framework for multi-robot coordination. Code and video: https://github.com/proroklab/d4orm

cs.RO↗
arXiv2026

Gel-Confined Rolling-Circle Amplification Enables Sensitive Single-Cell Proteoform Analysis

Protein abundance alone does not capture the molecular diversity generated by protein processing and post-translational modification, yet most single-cell protein assays do not resolve these proteoform states. Here, we develop RCAmp-scWB, a post-separation amplification strategy that performs rolling-circle amplification directly within the polyacrylamide gel after single-cell protein electrophoresis. Reaction-transport modeling identifies a balance between reagent access and confinement of the template-associated amplification product that supports signal amplification while retaining the electrophoretically encoded spatial readout. Using purified protein standards, RCAmp-scWB produced a 4.0-7.8-fold steeper concentration-response slope than conventional single-cell western blotting. Across six human cancer cell lines, RCAmp-scWB quantified protein abundance and resolved distinct Vimentin and PD-L1 proteoform states whose relative abundance and single-cell distributions were not captured by total protein measurements alone. Extension to individual small extracellular vesicles further revealed source-dependent shifts in Vimentin proteoform composition despite comparatively similar abundance of one major Vimentin species. By separating electrophoretic molecular discrimination from signal amplification, RCAmp-scWB enables proteoform-resolved analysis of heterogeneous cells and small extracellular vesicles.

q-bio.QM↗
arXiv2026

ScaRF-SLAM: Scale-Consistent Reconstruction with Feed-Forward Models and Classical Visual SLAM

Recent works have explored unifying SLAM with geometric foundation models (GFMs). However, directly using GFM predictions for tracking is highly sensitive to model capability and uncertainty, as geometric inaccuracies in the predictions can adversely affect pose estimation. To address this limitation, we propose a decoupled framework that integrates classical feature-based SLAM with GFMs, which achieves higher quality and more consistent dense reconstruction. In brief, we use classical visual SLAM for robust low-latency tracking and use GFMs exclusively for mapping. By anchoring mapping to poses produced by the SLAM module and optimizing across depth scales, the proposed design avoids propagating inaccuracies from GFM predictions into pose estimation while imposing geometric constraints on the reconstruction. The system builds submaps from multiple posed keyframes and enforces scale consistency via lightweight frame and submap scale optimization. It also performs projection-based point cloud fusion within each submap, and updates submaps online to reflect trajectory updates from the feature-based SLAM. To evaluate tracking and reconstruction of our method, we introduce a loop-rich, building-scale indoor dataset with accurate sensor trajectories and LiDAR ground-truth. Experiments show that our approach achieves superior trajectory accuracy while improving reconstruction precision by 10%-20% over existing methods, with about 2 cm reconstruction error per 10 m chunk on building-scale dataset. On large-scale outdoor datasets, it attains 10 cm error per 30 m chunk (w.r.t LiDAR ground-truth models). Code and dataset: https://github.com/ori-drs/ScaRF-SLAM

cs.RO↗
arXiv2026

SWE-Proof: Can Language Models Resolve Real-World Issues with Machine-Checked Proofs?

Ensuring the correctness of LLM-generated code is a core challenge for modern software engineering. Benchmarks for agentic code generation check correctness with held-out test suites, which are inherently incomplete and increasingly susceptible to memorization. Formal verification avoids both problems, but existing work covers only standalone tasks whose specifications are given as input, not real issues, which touch large repositories and state intent in vague natural language. We present Benchproofer, a pipeline that turns a coding task with a known correct patch into a formally verified one: it writes a specification for the new code, summarizes the existing functions that code calls with axioms, and admits an instance only after mechanical and adversarial gates agree. Applying it to SWE-bench Verified yields SWE-Proof, 500 real issues whose correctness is formally verified rather than tested, and it extends to SWE-bench Pro. Evaluating Claude Opus 4.8, we find that verification catches what tests miss: a quarter of test-passing patches admit counterexamples, which a structured natural-language specification does not fix, while a correct formal one lifts resolution from 85% to 95%. Writing that specification is the hard part: an agent that must write its own gains nothing over an unaided baseline, and only 56% of those specifications pass our audit. The usual failure is faithfulness, a specification that constrains part of the required behavior and leaves the rest free. Specification quality still tracks the outcome, failing on 92% of unresolved instances against 51% of resolved ones, making faithful specification synthesis a concrete open problem.

cs.LG↗
arXiv2026

M2Tok: Multi-head Multi-codebook Discrete Action Tokenization for Vision-Language-Action Models

Recent advancements have successfully adapted autoregressive language models to process multimodal signals, such as images and actions. Since raw action signals are continuous, effective tokenization is essential to map high-dimensional inputs into compact discrete tokens for autoregressive processing. However, existing discrete action tokenizers often suffer from high reconstruction loss, failing to preserve the fine-grained dynamics required for precise control. This "discretization bottleneck" significantly limits the performance ceiling of downstream Vision-Language-Action (VLA) models. To address this, we propose ${M}^2$Tok, a Multi-head Multi-codebook Action Tokenizer designed to minimize reconstruction error and enhance policy performance. Our approach introduces two key structural innovations: (1) we decompose the latent action features into multiple heads, enabling the model to implicitly align specific heads with distinct action dimensions; (2) we assign independent codebooks to each head for quantization. By leveraging the combinatorial nature of multiple codebooks, we significantly expand the representational expressivity of the tokenizer, leading to substantially lower reconstruction loss compared to previous methods. We evaluate the ${M}^2$Tok-based VLA on the RoboTwin, Simpler-Env, and 3 zero-shot real-world tasks. Experimental results demonstrate our method not only achieves superior reconstruction fidelity but also significantly boosts the success rate of VLA models. Comprehensive ablation studies further confirm the effectiveness of the multi-head and multi-codebook mechanisms. Code is available at https://github.com/cpaaax/M2Tok.

cs.RO↗