Search arXiv⌕ Search

arXiv subjects

Yuan Yao

Publications and source records attributed to Yuan Yao.

At least 19 recordsLinked to original sources
arXiv2026

RoboDoctor: Training-Free Recursive Self-Improvement for Robot Policies

Generalist robot policies and language-model agents provide complementary capabilities, but completing a task also requires adapting their use to the consequences of physical interaction. We study execution as a locus of adaptation: feedback guides the current action and supplies evidence for revising subsequent decisions under fixed model weights. We introduce RoboDoctor, a training-free framework that connects diagnosis, grounded action, and persistent experience across three revision targets. The harness allocates control between policy and agent; intervention determines when to assist while preserving useful policy behavior; and operation refines physical methods through observed effects. Linked experience records retain the reason for a decision, the attempted method, its outcome, and conditions for reuse. On a curated archive of 54 RoboDojo-Sim task entries, RoboDoctor achieves a five-dimension mean score of 57.10, with a native success rate of 49.44%. Recorded interactions illustrate strategy changes, selective correction, and grasp revision. These findings characterize execution-time adaptation in the evaluated system; controlled trials remain necessary to isolate the contribution of cross-episode experience.

cs.RO↗
arXiv2026

Mobile Backscatter Communication for the Battery-less Internet of Things

We enable backscatter communication in the battery-less mobile Internet of Things (IoT). Backscatter communication is extensively studied in static settings. Existing designs are, however, fundamentally mismatched with mobility and time-varying energy patterns. Channel conditions rapidly fluctuate, impacting the achievable data rates and thus transmission costs. Energy availability varies unpredictably, possibly forcing devices to remain quiescent to recharge energy buffers. The two issues compound each other: while recharging, a battery-less mobile IoT device may miss more favorable channel conditions. We design a lightweight decision system that dynamically determines when to transmit by checking short-term trends in signal strength, while using Non-volatile Memory (NVM) to retain packets in unfavorable channel conditions and across energy failures. Using a prototype we built and real-world mobility and power traces, we compare our design against a rate-adaptive baseline that only considers the instantaneous channel conditions. Experimental results show that our system improves throughput by up to 5.16x while reducing transmission energy consumption by up to 47.3%, with only 0.23% - 7.3% additional energy overhead.

cs.NI↗
arXiv2026

RAIN: Agentic Proof Generation for Rocq-based Software Verification

Interactive theorem provers can provide strong correctness guarantees for software, but constructing the proofs remains a challenging and labor-intensive task. Inspired by the recent progress of modern coding agents, we present RAIN, a lightweight proof-specialized agent for Rocq proof generation. RAIN augments an LLM with proof engineering capabilities, including repository exploration, iterative refinement, and checker feedback. This design allows the agent to retrieve lemmas autonomously, revise complete proof scripts instead of isolated tactics, and iteratively refine its proof strategy using enriched proof-state information. Evaluated on a subset from CoqStoq, RAIN shows substantial improvements over the existing baselines, achieving over 90% prove success rate with at least 24.7% improvements. We also compare with general-purpose coding agents, the results show that RAIN maintains comparable prove success rate but with significant lower cost.

cs.LO↗
arXiv2026

What Should Robot Policies Learn to Predict? Rethinking Supervision Targets in Vision-Language-Action Learning

Vision-language-action (VLA) policies are typically trained with action prediction as the primary supervision signal, since robots ultimately execute actions. Auxiliary objectives such as latent-action (LA) and future-vision (FV) prediction are often added to improve policy learning, yet they compete with action prediction for the same training budget. We ask whether, under a fixed budget, allocating a substantial share of training samples to such objectives, rather than to further action prediction, can improve policy learning. We build a unified VLA framework that couples a two-stream multimodal diffusion transformer (MMDiT) with a sample-level router, which assigns each training sample to either action prediction or an auxiliary objective and thus makes supervision allocation explicit. Within this framework, we instantiate three LA targets based on LAPA and its DINO-feature variants, and three FV targets based on DINOv3, V-JEPA 2, and Wan2.2 representations. Across LIBERO, CALVIN, and four real-robot tasks, suitable LA and FV supervision outperforms action-only training under the same budget, even when action prediction receives no more than half of the training samples. FV generally outperforms LA, although the strongest LA target remains competitive on CALVIN. Diagnostic analyses suggest that DINO-based LA targets emphasize relative motion, whereas FV targets emphasize absolute position and task identity. We therefore argue that what robot policies learn to predict, and how training is allocated among these targets, should be treated as explicit design choices rather than defaulting to action-dominant training.

cs.RO↗
arXiv2026

Low-cost algorithm-to-execution framework for surface-code quantum computing

Fault-tolerant quantum computing is progressing from demonstrations of individual logical components toward the execution of useful quantum algorithms on large-scale error-corrected processors. Realizing this transition requires co-design across algorithms, fault-tolerant implementations, and hardware resources. Here we develop an algorithm-to-execution framework for low-cost surface-code quantum computing that carries the structure of a computation from hierarchical logical descriptions to circuit-specific spatial organization, non-Clifford resource planning, and physical resource evaluation. We apply the framework to twenty benchmark circuits spanning seven algorithm families and to a hierarchically reconstructed application-scale elliptic-curve discrete-logarithm workload. Under the stated error and resource-supply models, combining circuit-specific layouts with hybrid non-Clifford implementation reduces space--time volume by up to 12.70x relative to an all-synthesis Litinski fast-block baseline. Circuit-specific layouts reduce both modeled runtime and space--time volume in eighteen benchmarks, while the remaining two retain the fast layout. These results show that the physical cost of fault-tolerant computation emerges from the interaction between algorithm structure, error-corrected implementation, and hardware organization rather than from aggregate logical resource counts alone.

quant-ph↗
arXiv2026

PVFH-Bench: Benchmarking Premise-Induced Visual Fixation Hallucinations in Vision-Language Models

Large Vision-Language Models (LVLMs) have achieved remarkable progress across a wide range of multimodal tasks. However, object hallucination remains a persistent challenge, undermining model reliability in applications such as visual question answering, embodied agents, and human-AI decision support systems. In this paper, inspired by cognitive science studies on perceptual bias, we investigate a new failure mode termed Premise-Induced Visual Fixation Hallucination (PVFH). Specifically, PVFH refers to a premise-induced reasoning error in which models become anchored to visually confusable evidence, accept a false textual premise, and subsequently generate responses about an absent yet visually plausible object. To systematically study this phenomenon, we introduce PVFH-Bench, a benchmark that automatically constructs visually confusable but absent targets and pairs the core hallucination setting with diagnostic subsets to trace how models move from object recognition to Premise-Induced Visual Fixation Hallucination. Experimental results on 10 state-of-the-art LVLMs demonstrate that all evaluated models are highly vulnerable to Premise-Induced Visual Fixation Hallucination. Moreover, we find that explicit chain-of-thought reasoning can further exacerbate this failure mode by encouraging models to overinterpret weak visual cues as supporting evidence for hallucinated objects.

cs.CL↗
arXiv2026

Optical Signals Synthesized from an Optical Lattice Clock with Uncertainty of 1.3E-17

We report an accurate optical frequency synthesizer, which can generate single-frequency laser light with high frequency stability and accuracy at desired frequencies over a wide optical region. The frequency of the output signal is divided from an 171Yb optical lattice clock via an accurate optical frequency divider based on an optical frequency comb. Therefore, the output of the optical frequency synthesizer inherits the frequency accuracy from the Yb optical clock. The frequency uncertainty of the 171Yb optical lattice clock is evaluated to be 6.8 mHz, corresponding to a fractional frequency uncertainty of 1.3E-17, mainly limited by the blackbody radiation shift and the lattice-induced light shift.

physics.optics↗
arXiv2026

Accurate recovery of the two linewidths hidden in laser beatnote statistics

Photons from ultrastable lasers can remain coherent over distances approaching the Earth-Sun separation, enabling optical clocks projected to lose less than one second over the age of the Universe. Yet characterizing these photons poses an identifiability problem: a two-laser heterodyne measurement produces a single beatnote linewidth that conflates contributions from both lasers. For more than half a century, the standard solution has been the three-cornered-hat (TCH) method, which requires three independent ultrastable lasers. Here we derive the mathematical form and elucidate the physical origin of asymmetric beatnote-linewidth distributions arising from finite photon wave trains, a long-observed feature not captured by canonical Gaussian or Lorentzian statistics. This finding accurately recovers two linewidths hidden in laser beatnote statistics without a third laser. The framework consistently captures the observed coherence-length and coherence-time statistics of photons, while comparison with TCH measurements confirms the quantitative validity of the extracted individual linewidths across five independent ultrastable-laser datasets spanning nearly two orders of magnitude. Most notably, the method resolves the 7.8-mHz linewidth of a cryogenic silicon-cavity laser beating with a broader laser. This transformative function of beatnote-linewidth distributions, together with the resulting method, could fundamentally advance optical clocks and precision metrology.

physics.optics↗
arXiv2026

Circuit-level benchmarks of GKP-concatenated qLDPC Codes

Scalable fault-tolerant quantum error correction based on Gottesman-Kitaev-Preskill (GKP) codes requires finite-rate outer codes that can exploit analog GKP information under realistic circuit noise. However, quantum low-density parity-check (qLDPC) outer codes have not been systematically compared as candidates for circuit-level GKP concatenation. Here, we benchmark BB and tricycle codes using a unified simulation framework that progresses from code-capacity noise to repeated noisy syndrome extraction and schedule-resolved displacement propagation. Across all three noise models, analog-informed belief-propagation with ordered-statistics decoding (BP-OSD) yields higher finite-size crossing estimates than hard-decision decoding. In the full circuit model, the analog-informed crossings are \(σ_{\mathrm{th}}\simeq0.212\) for the selected BB sequence and \(0.142\) for the tricycle sequence, corresponding to squeezing requirements of approximately \(10.46\,\mathrm{dB}\) and \(13.94\,\mathrm{dB}\), respectively. These results show that analog GKP information continues to improve decoding under circuit-level noise. Under the common simulation assumptions used here, the selected BB sequence also yields a higher finite-size crossing estimate than the selected tricycle sequence.

quant-ph↗
arXiv2026

DRIFT: Data Selection for LLM Instruction Tuning via On-Policy Attribution

Data selection for instruction tuning is usually studied for efficiency, reaching strong performance with a fraction of the training data. We ask whether data selection can instead raise the performance ceiling, producing a better model than training on all the available data. To test this, we take a large language model (LLM) that has already completed supervised fine-tuning (SFT) on its full corpus and train it further on examples selected from that same corpus. Any gain therefore lifts the model above the ceiling set by full-data training. Existing selection methods yield little or no gain in this setting. We propose DRIFT (Data Refinement via On-Policy Influence Functions for Supervised Fine-Tuning), a data attribution method built on influence functions. DRIFT samples the model's own responses to validation queries and weights them by correctness to form a validation objective. It ranks the examples of the original corpus by their estimated influence on this objective, and the model then continues SFT on the top-ranked subset. This design is motivated by the proximity gap in influence function theory: we hypothesize that these on-policy responses are better validation targets than external references in terms of locality. DRIFT also corrects influence scores for their dependence on gradient norm within each validation task. DRIFT outperforms all evaluated data selection baselines on Olmo3-7B-Instruct-SFT and OpenR1-Distill-7B, raising average accuracy across eight benchmarks from 39.17 to 40.14 and from 57.23 to 58.66, respectively, with gains on benchmarks not used for attribution.

cs.LG↗
arXiv2026

SimpleMemVLA: A Simple but Effective Native-Video Memory for Vision-Language-Action Models

Long-horizon manipulation is partially observable: the information needed to choose the next action may appear only in observations from minutes earlier. Existing memory mechanisms for VLAs, such as retrieval banks, learned compressors and recurrent states, must decide what to keep from the past before knowing what a future decision will require. They were motivated by the assumption that minute-scale history is too large to process directly, which no longer holds for modern VLM backbones. We propose SimpleMemVLA, a VLA without a dedicated memory module that uses the backbone's native video context directly as memory. It keeps the sampled history intact in the timestamped video format the backbone was pretrained to process, routes the evidence it finds to a standard flow-matching action head through the hidden states of a generated sub-task, and prefills the history shared by consecutive decisions during action execution, keeping latency close to that of a single-frame VLA. SimpleMemVLA achieves state-of-the-art results on four memory benchmarks without loss on general-purpose control, and with the same backbone and training setup it outperforms retrieval, compression and recurrent-state methods by a wide margin. History interventions show that the policy reads specific evidence from its past and follows edited histories without parameter updates, a visual form of in-context learning. On a physical dual-arm robot, it completes two tasks whose decisive evidence disappears before the robot acts. Code available at https://github.com/OpenBMB/SimpleMemVLA

cs.CV↗
arXiv2026

DepthEvidence: Unifying Metric Depth Prediction and Geometric Reasoning in Multimodal Language Models

Spatial reasoning with metric constraints requires linking objects to geometric measurements and preserving their numerical content during language reasoning. We present DepthEvidence, a 4B model that uses its own dense metric predictions as object-grounded evidence for language generation. A camera-conditioned decoder predicts full-resolution metric depth using multi-scale visual features and high-resolution RGB refinement. A dense-to-language interface converts predicted depths and decoder features into object-aligned continuous geometry tokens anchored to object identifiers. Geometric supervision encourages metric information to remain recoverable before and after language-context interaction, while instruction tuning supports object measurement and compositional reasoning. We introduce a Depth-VQA benchmark evaluating object-depth queries, relative comparisons, and decisions combining spatial and numerical constraints. Across nine datasets, DepthEvidence achieves the highest average dense $δ_1$ among evaluated methods, competitive with specialized estimators. It also leads the evaluated methods in instance-level metric depth estimation and overall accuracy on both relative and metric reasoning tracks, while broadly preserving general VQA performance and improving spatial understanding relative to the base model.

cs.CV↗
arXiv2026

Unimodality of Independence Polynomials for Sufficiently Large Forests

We prove that the independence sequence of every sufficiently large forest is unimodal. The result follows from establishing log-concavity on a central interval of the sequence, along with monotonicity of the initial and final segments. The main analytic step is a central limit theorem for the size of a random independent set sampled with the hard-core model, uniform over all forests and over an interval of positive fugacities.

math.CO↗
arXiv2026

Liberating LLM Capabilities in Full-Duplex Speech Models

Speech-based large language models are typically constrained to spoken replies, which limits their user-facing outputs to what can be verbalized and suppresses text-native capabilities such as code generation, structured analysis, and multi-step reasoning in realtime interaction, for tasks that require persistent, structured, and inspectable intermediate outputs. Existing work improves spoken reasoning or full-duplex turn-taking, but still treats text as a hidden intermediate state or a subordinate modality rather than a first-class output channel. We propose Listen-Write-Speak (LWS), a text-first tri-channel paradigm in which a single autoregressive LLM continuously listens to user audio, writes visible free-form text as its primary output, and speaks a realtime oral response in parallel under a shared causal attention context. This behavior is implemented entirely through a Token Schema, requiring no architectural modifications, and learned via a two-stage data pipeline that synthesizes per-second cognitive annotations consistent with the revealed input timeline. Empirically, LWS demonstrates strong full-duplex interaction on Full-Duplex-Bench, reaches 4.72 on VoiceBench AlpacaEval, achieves 92.6% writing-speaking consistency, and consistently outperforms its internal ablations on URO-Bench. These results suggest that visible writing can serve as a first-class output channel for speech interaction without sacrificing realtime responsiveness. The code and dataset are available on the project page: https://royalzhang.com/project/lws-page/.

cs.CL↗
arXiv2026

Extremely Low-Cost Magic State Preparation toward Fault-Tolerant Quantum Computing

Fault-tolerant preparation of non-Clifford resource states is a major contributor to the overhead of quantum computation, motivating protocols that achieve high output fidelity with minimal qubit and circuit costs. We introduce a low-cost magic-state preparation protocol in which the choice of stabilizer generators is co-designed with the flag gadgets, allowing the syndrome-extraction circuit itself to filter correlated faults across a non-Clifford layer. The protocol prepares a logical plus state in the 15-qubit quantum Reed-Muller code, applies a transversal T gate, and gauge-fixes the same register into the seven-qubit Steane code. By reorganizing equivalent Z-type stabilizer generators into jointly flagged measurement groups, the protocol eliminates all accepted logical-error contributions arising from one or two circuit faults under destructive error detection. Under a uniform circuit-level depolarizing noise model, the postselected infidelity is $210.2p^3+O(p^4)$. At $p=10^{-3}$, exact low-order enumeration combined with stratified sampling bounds the infidelity by $2.2\times10^{-7}$ at 99.9% joint confidence, while retaining an acceptance probability of 86.9%. The complete circuit requires only 19 qubits and 82 CNOT gates. These results demonstrate that stabilizer-generator design can substantially reduce the cost of postselected magic-state preparation, although corrected operation and the fidelity of an unmeasured output block require separate analysis.

quant-ph↗
arXiv2026

Binary X-rays of doubly stochastic matrices

The X-ray of a permutation is a sequence of sums along each diagonal of the associated permutation matrix. They satisfy certain necessary constraints on distribution of the values, which are conjectured to be sufficient when the sequence is binary. By re-expressing the constraints in a form that allows for real-valued relaxations, we prove that these binary sequences are always X-rays of doubly stochastic matrices.

math.CO↗
arXiv2026

Large Exchange Magnetostriction in a Kagome Antiferromagnet at Room Temperature

The pursuit of high-performance magnetostrictive materials is crucial for advancing technologies in sensing, actuation, and microelectromechanical systems. Although pronounced magnetostrictive effects have been observed in a few ferromagnets, a systematic exploration of magnetostriction across a broader range of antiferromagnets remains limited. Here, we report the observation of large magnetostriction in the kagome antiferromagnet YMn6Sn6. Under a magnetic field, the system undergoes a field-induced evolution from a helical magnetic ground state toward a collinear field-polarized state, accompanied by a large anisotropic lattice strain and a substantial volume magnetostriction exceeding 400 ppm at room temperature. Remarkably, the magnetostrictive response remains nearly fully reversible up to 9 T with nearly hysteresis-free behavior, effectively minimizing the energy dissipation commonly associated with domain-wall pinning. Combined experimental measurements and theoretical calculations reveal that the large, nearly hysteresis-free magnetostriction originates from the competition between intralayer ferromagnetic and interlayer exchange interactions. This exchange-driven magnetoelastic coupling further gives rise to a strongly direction-dependent lattice distortion pathway. This work establishes kagome helimagnets as a tunable platform for low-dissipation magnetoelastic functionalities and responsive magnetomechanical applications.

cond-mat.mtrl-sci↗
arXiv2026

Symmetry-Enforced Topological Structures in Quantum Phase Diagrams

We study the topological structure of the quantum phase diagram of gapped systems by identifying the noncontractibility of loops, so-called $S^1$-families, within the gapped phase diagram in which many-body Hamiltonians can have nontrivial ground-state degeneracy. We manifest the role of symmetries in such $S^1$-family classifications by the exotic symmetry interplay: (i) nontrivial mixed anomalies, (ii) semi-direct product relation between the spontaneously broken and the unbroken symmetries, and (iii) symmetry with noninvertible operators. We find that such structures lead to the novel $S^1$-family classifications inaccessible by earlier classifications based on ``independent'' symmetries. Furthermore, we construct lattice realizations of these $S^1$-families and explicitly demonstrate their novel algebraic structures.

cond-mat.str-el↗