arXiv subjects
Publications and source records attributed to Yiwei Liu.
In this paper, based on the warped product structures determined by the Obata-type equations with Robin boundary condition, we establish some rigidity results for compact manifolds with smooth boundary under appropriate Bakry-\'Emery Ricci curvature conditions and some other assumptions.
Natural-language instruction changes can directly alter robot behavior. A reliable embodied system should preserve its action when the task is unchanged and update it correctly when the task itself changes. We introduce One Word, Different Action, a real-robot benchmark built on physical decision states and executable actions, using task-preserving and task-changing instruction pairs to jointly evaluate Decision Invariance and Decision Sensitivity, with further evaluation under multi-constraint reasoning and real-RGB grounding. Experiments show that modern models are near saturation on single-constraint instruction changes, yet several models degrade noticeably when multiple task constraints must be integrated into one executable decision. These results suggest that the more salient remaining challenge is no longer recognizing an isolated instruction change, but reliably composing multiple task requirements into a correct robot action decision.
This paper documents a frozen engineering project on personalized Chinese handwriting. The project started from approximately 200 real handwriting images from one user, covering 197 unique Chinese characters, and was initially formulated as few-shot generation of unseen characters. A sequence of canonical-centered personalization routes repeatedly exposed the same conflict: increasing structural pressure made outputs more canonical, while increasing personalization could damage identity-defining strokes. The project was therefore reset around real-human character equivalence classes. A multi-writer CASIA candidate pool showed that a USER-compatible realization often already existed among valid human samples. The task consequently changed from synthesis to character-wise matching, followed by cross-writer composition into a virtual writer. The frozen system uses real-ink features, character-specific human population percentiles, top-20 candidate pruning, and greedy hardest-first whole-row selection. On the covered target set, all 197 USER characters had real-human candidates, and the 100-character evaluation subset was covered 100/100. Knowncharacter held-out comparisons included a row judged visually almost indistinguishable from genuine USER handwriting. A 60- episode stability audit placed every episode in a predefined A-like machine-proxy region, but these were not independent human A-level judgments. The final evidence supports stable practical B-level quality, with many outputs approaching A-level under the USER-defined criterion. The report records why generation became unnecessary for this case without claiming unrestricted or universal handwriting synthesis.
Natural language provides robots with a flexible task interface, but target ambiguity in embodied environments arises not only from user intent; it can also result from missing taskrelevant physical evidence in the current observation. Existing interactive disambiguation methods primarily obtain additional information by asking the user, whereas occlusion, restricted viewpoints, unreadable text, and unobserved targets require the robot to actively change its observation. We propose an active-perception framework for embodied target disambiguation that uses active observation as the backbone for information acquisition and uses a vision-language model to decide, on the basis of accumulated visual evidence and interaction information, whether to continue observing, request clarification, or complete target selection. Active observation can both directly recover missing discriminative evidence and reveal object names, labels, and semantic attributes, thereby improving user clarification when it remains necessary. Real-robot experiments show that the framework combines physical information acquisition and userintent clarification within a unified embodied disambiguation process.
Enabling large language models to achieve stable self-improvement without external expert supervision remains a central challenge in complex reasoning tasks. Existing self-distillation and reinforcement learning methods lack explicit mechanisms for tracking problem-level learning progress and adapting optimization strategies accordingly. Consequently, training may over-optimize easy problems, receive weak supervision from hard problems, and fail to sufficiently explore borderline cases. To resolve these issues, we propose DRIFT, an online self-evolution policy optimization framework for large language models. DRIFT regulates the model's self-improvement process through the joint use of Difficulty Routing and Rhythm Gating. The former identifies the model's learning state at the problem level and dynamically allocates self-distillation and reinforcement learning signals, while the latter refines policy updates at the token level, concentrating exploration on critical reasoning positions. By further incorporating a success buffer and a two-stage curriculum learning strategy, DRIFT preserves high-quality historical experience while progressively guiding the model from reliable behavior acquisition toward stable policy evolution. Evaluated across five benchmarks and three model scales, DRIFT surpasses the peak performance of both GRPO and SDPO across all evaluated metrics. On the average score over the five benchmarks, DRIFT achieves 79.5$\%$, outperforming GRPO by 9.5$\%$ and SDPO by 7.5$\%$, establishing a new state-of-the-art result. Notably, on ToolUse, DRIFT reaches an accuracy of 79.2$\%$, improving over GRPO by 13.5$\%$ and SDPO by 10.7$\%$, setting a new state-of-the-art and substantially outperforming all concurrent methods.
Recent years have witnessed the rapid evolution of AI agents toward handling increasingly complex, real-world tasks. However, existing benchmarks rarely evaluate whether agents can operate graphical user interfaces to complete long-horizon, high-value professional workflows across diverse domains. Current GUI benchmarks still predominantly focus on general-purpose software, relatively simple applications, and short-horizon tasks, leaving it largely unknown whether modern agents can follow user instructions to autonomously operate domain-specific professional software and accomplish economically valuable work in an end-to-end manner. To bridge this gap, we introduce Workflow-GYM, a benchmark for long-horizon GUI tasks centered on professional domains and specialized software environments. Through extensive experiments on state-of-the-art models, we find that even the strongest models achieve only slightly above 30% success rates, highlighting that professional long-horizon GUI workflows remain highly challenging for current GUI agents. Further analysis reveals that current agents struggle to maintain long-horizon workflow consistency, frequently exhibiting workflow stage omission, error propagation, objective drift, and insufficient understanding of professional software environments. Our findings provide important insights into the limitations of current agent systems and suggest key directions for the next generation of GUI-agent research.
We study \textsc{OCO-S$^2$}, an online convex optimization setting in which decisions drive a stable dynamical state, losses are incurred along the induced state trajectory, and first-order feedback is available only through sparse block communication with partial participation. This coupling creates a dynamic-regret problem beyond pointwise OCO: the learner updates and holds decisions at the block scale, whereas the hindsight comparator may vary at the per-round scale. We propose \textsc{OCO-S$^2$-OGD}, a projected block online gradient method that updates deployed decisions using sparse block-level distributed feedback. We prove dynamic-regret bounds for the incurred trajectory cost, quantifying the tradeoff among block communication, comparator variation, state-memory truncation, and partial participation. We further introduce a prediction-augmented variant, \textsc{OCO-S$^2$-OGD-P}, and show that accurate block-level predictions improve the optimization term in the regret bound through their realized gradient-mismatch error. Overall, this work provides a regret-theoretic foundation for communication-efficient online decision-making in systems where algorithmic updates and physical state trajectories are intrinsically coupled.
This paper studies secondary frequency control in transmission networks subject to communication delays at the cyber-physical interface and limited per-update computation at the control center. The regulation objective is formulated as a constrained economic dispatch problem incorporating generation capacity constraints, nodal power balance, transmission-flow limits, and scheduled tie-line power exchanges. Based on this formulation, we develop a passivity-based control framework in which an augmented projected primal-dual controller restores nominal frequency and drives the closed-loop system to the solution set of the constrained economic dispatch problem. Two-way communication delays between the physical network and the control center are modeled as scattering-based passive channels for the measurement uplink and the control-command downlink. This construction preserves the target equilibrium and enables a delay-robust passivity analysis of the delayed closed loop. To reduce the computational burden at the control center, we develop a randomized block-coordinate implementation of the augmented projected primal-dual controller. The resulting sampled-data closed loop preserves the target solution set and achieves local mean-square geometric convergence under suitable step-size and regularity conditions. Finally, a multivariable wave-domain interface filter is introduced to inject additional dissipation and improve the damping of the delayed interface without altering the steady-state interconnection. Simulations on the IEEE 14-bus system indicate that the proposed digital implementation accurately reproduces the delayed closed-loop behavior while reducing the per-update computational cost.
Automating radio frequency (RF) amplifier design remains challenging because existing methods suffer from the curse of dimensionality, weak use of domain knowledge, and poor transferability, leading to low data efficiency. Meanwhile, although large language models (LLMs) have shown promise in many scientific domains, applying them directly to RF sizing is nontrivial due to the numerical nature of circuit optimization and the reliance on domain-specific design flows. To address this, this paper proposes RFAmpDesigner, a multi-agent framework that automates RF amplifier sizing. It introduces a resource-allocation middleware that reframes high-dimensional parameter tuning as a low-dimensional resource distribution problem, making it easier to inject sizing knowledge into general-purpose LLMs. The framework also follows standard design practice, enabling LLMs to distinguish between high- and low-cost actions and search in parallel. To realize a self-evolving optimization process, the framework employs retrieval-augmented generation (RAG) to reuse past knowledge and experience from memory base. As a proof of concept, we apply RFAmpDesigner to low noise amplifiers of varying complexity. The experimental results show that it can automatically synthesize designs with fractional bandwidths ranging from 10\% to 80\% and center frequencies from 10 GHz to 50 GHz. To the best of our knowledge, this work develops the first LLM-driven approach for RF amplifier sizing that operates on design concepts instead of treating netlists as text, offering a novel solution to mitigate data scarcity in RF design.
Bimanual manipulation is imperative yet challenging for robots to execute complex tasks, requiring coordinated collaboration between two arms. However, existing methods for bimanual manipulation often rely on costly data collection and training, struggling to generalize to unseen objects in novel categories efficiently. In this paper, we present Bi-Adapt, a novel framework designed for efficient generalization for bimanual manipulation via semantic correspondence. Bi-Adapt achieves cross-category affordance mapping by leveraging the strong capability of vision foundation models. Fine-tuning with restricted data on novel categories, Bi-Adapt exhibits notable generalization to out-of-category objects in a zero-shot manner. Extensive experiments conducted in both simulation and real-world environments validate the effectiveness of our approach and demonstrate its high efficiency, achieving a high success rate on different benchmark tasks across novel categories with limited data. Project website: https://biadapt-project.github.io/
Let $(\Omega^{n+1}, g)$ be an $(n+1)$-dimensional smooth compact connected Riemannian manifold with smooth boundary $\Sigma$, satisfying that ${\text{Ric}_{\Omega}}\ge 0$ and $\Sigma$ is strictly convex, more precisely, its second fundamental form $h\ge cg_{\Sigma}$ for some positive constant $c$. Escobar {\cite{escobar1997geometry}} considered the first nonzero Steklov eigenvalue $\sigma_1$ of $(\Omega^{n+1}, g)$ and proved that $\sigma_1\geq c$ when $n=1$ and $\sigma_1>{\frac{c}{2}}$ when $n \geq 2$. He then conjectured {\cite{escobar1999isoperimetric}} that the first nonzero Steklov eigenvalue $\sigma_1\ge c$. Very recently, Xia and Xiong {\cite{xia2023escobar}} confirmed Escobar's conjecture in the case that $\Omega$ has nonnegative sectional curvature, by constructing a weight function and using appropriate integral identities. In this paper, we construct a new weight function under certain sectional curvature assumptions and provide some new lower bounds for the first nonzero Steklov eigenvalue, which can be considered as generalizations of the results of Escobar and Xia-Xiong. As an application of the weight function, we also consider lower bound estimate of the first nonzero Steklov eigenvalue under conformal transformations.
Drug synergy prediction is a critical task in the development of effective combination therapies for complex diseases, including cancer. Although existing methods have shown promising results, they often operate as black-box predictors that rely predominantly on statistical correlations between drug characteristics and results. To address this limitation, we propose CausalDDS, a novel framework that disentangles drug molecules into causal and spurious substructures, utilizing the causal substructure representations for predicting drug synergy. By focusing on causal sub-structures, CausalDDS effectively mitigates the impact of redundant features introduced by spurious substructures, enhancing the accuracy and interpretability of the model. In addition, CausalDDS employs a conditional intervention mechanism, where interventions are conditioned on paired molecular structures, and introduces a novel optimization objective guided by the principles of sufficiency and independence. Extensive experiments demonstrate that our method outperforms baseline models, particularly in cold start and out-of-distribution settings. Besides, CausalDDS effectively identifies key substructures underlying drug synergy, providing clear insights into how drug combinations work at the molecular level. These results underscore the potential of CausalDDS as a practical tool for predicting drug synergy and facilitating drug discovery.
Joint logical-numerical reasoning remains a major challenge for language models, yet existing datasets rely on fixed rule sets and offer limited control over task complexity, constraining their generalizability for evaluation and training. We present LogiNumSynth, a flexible natural language problem synthesizer that synthesizes tasks requiring proficiency in joint logical reasoning (e.g., rule-based reasoning) and numerical reasoning (e.g., arithmetic computation). LogiNumSynth supports fine-grained control over reasoning world richness, logical reasoning depth, and the complexity of numerical computations, enabling flexible data synthesis across difficulty levels. We demonstrate three key contributions: (1) Synthesizer -- synthesizing fully controllable joint reasoning tasks over natural language; (2) Evaluation & Process Analysis -- evaluating both process accuracy and answer accuracy; (3) Targeted Training -- using synthesized data to enhance LLMs' reasoning performance. Experiments with multiple LLMs highlight persistent weaknesses in logical-numerical reasoning, showing that LogiNumSynth can serve as both a diagnostic tool and a source of targeted supervision for advancing integrated reasoning skills.
While recent studies explore Large Language Models' (LLMs) performance on Theory of Mind (ToM) reasoning tasks, research on ToM abilities that require more nuanced social context is limited, such as white lies. We introduce TactfulToM, a novel English benchmark designed to evaluate LLMs' ability to understand white lies within real-life conversations and reason about prosocial motivations behind them, particularly when they are used to spare others' feelings and maintain social harmony. Our benchmark is generated through a multi-stage human-in-the-loop pipeline where LLMs expand manually designed seed stories into conversations to maintain the information asymmetry between participants necessary for authentic white lies. We show that TactfulToM is challenging for state-of-the-art models, which perform substantially below humans, revealing shortcomings in their ability to fully comprehend the ToM reasoning that enables true understanding of white lies.
Graphical User Interface (GUI) Agents, powered by large language and vision-language models, hold promise for enabling end-to-end automation in digital environments. However, their progress is fundamentally constrained by the scarcity of scalable, high-quality trajectory data. Existing data collection strategies either rely on costly and inconsistent manual annotations or on synthetic generation methods that trade off between diversity and meaningful task coverage. To bridge this gap, we present GUI-ReWalk: a reasoning-enhanced, multi-stage framework for synthesizing realistic and diverse GUI trajectories. GUI-ReWalk begins with a stochastic exploration phase that emulates human trial-and-error behaviors, and progressively transitions into a reasoning-guided phase where inferred goals drive coherent and purposeful interactions. Moreover, it supports multi-stride task generation, enabling the construction of long-horizon workflows across multiple applications. By combining randomness for diversity with goal-aware reasoning for structure, GUI-ReWalk produces data that better reflects the intent-aware, adaptive nature of human-computer interaction. We further train Qwen2.5-VL-7B on the GUI-ReWalk dataset and evaluate it across multiple benchmarks, including Screenspot-Pro, OSWorld-G, UI-Vision, AndroidControl, and GUI-Odyssey. Results demonstrate that GUI-ReWalk enables superior coverage of diverse interaction flows, higher trajectory entropy, and more realistic user intent. These findings establish GUI-ReWalk as a scalable and data-efficient framework for advancing GUI agent research and enabling robust real-world automation.
This paper addresses the prescribed performance control (PPC) challenge for high-order nonlinear systems affected by mismatched disturbances. The research aims to prevent singularity issues arising from error boundary violations during abrupt changes in reference trajectories. We introduce a novel transformation function with infinite-order differentiability at connection points, advancing beyond mere continuous differentiability. Utilizing this transformation function, we develop a comprehensive transformation strategy that ensures: (1) errors remain within prescribed boundaries when reference trajectories are smooth, and (2) errors return to prescribed boundaries within a specified timeframe following abrupt changes in reference trajectories. Additionally, the complexity explosion issue inherent in backstepping design is effectively resolved. Simulation results corroborate the validity of the proposed theoretical advancements.
This paper explores the voltage regulation challenges in boost converter systems, which are critical components in power electronics due to their ability to step up voltage levels efficiently. The proposed control algorithm ensures fixed-time stability, a desirable property that guarantees system stability within a fixed time frame regardless of initial conditions. To tackle the common chattering issues in conventional fixed-time control methods, a novel class of function families is introduced. State observers and adaptive parameters are utilized to manage the uncertainties associated with unknown load resistance. Furthermore, a new disturbance observer is developed using the proposed function family, and its advantages and limitations are illustrated through comparison with existing designs. Finally, both non-real-time and real-time simulations are conducted to validate the effectiveness and deployability of the proposed control algorithm.
We propose a theoretical framework for consistent channel hopping algorithms to address the multichannel rendezvous problem (MRP) in wireless networks with heterogeneous available channel sets. A channel selection function is called consistent if the selected channel remains unchanged when the available channel set shrinks, provided the selected channel is still available. We show that all consistent channel selection functions are equivalent to the function that always selects the smallest-index channel under appropriate channel relabeling. This leads to a natural representation of a consistent channel hopping algorithm as a sequence of permutations. For the two-user MRP, we characterize rendezvous time slots using a fictitious user and derive tight bounds on the maximum time-to-rendezvous (MTTR) and expected time-to-rendezvous (ETTR). Notably, the ETTR is shown to be the inverse of the Jaccard index when permutations are randomly selected. We also prove that consistent channel hopping algorithms maximize the rendezvous probability. To reduce implementation complexity, we propose the modulo algorithm, which uses modular arithmetic with one-cycle permutations and achieves performance comparable to locality-sensitive hashing (LSH)-based algorithms. The framework is extended to multiple users, with novel strategies such as stick-together, spread-out, and a hybrid method that accelerates rendezvous in both synchronous and asynchronous settings. Simulation results confirm the effectiveness and scalability of the proposed algorithms.