arXiv subjects

Yuheng Wu

Publications and source records attributed to Yuheng Wu.

At least 37 records · Page 2Linked to original sources

arXiv2026

NaviNote: Enabling In-situ Spatial Annotation Authoring to Support Exploration and Navigation for Blind and Low Vision People

GPS and smartphones enable users to place location-based annotations, capturing rich environmental context. Previous research demonstrates that blind and low vision (BLV) people can use annotations to explore unfamiliar areas. However, current commercial systems allowing BLV users to create annotations have never been evaluated, and current GPS-based systems can deviate several meters. Motivated by high-accuracy visual positioning technology, we first conducted a formative study with 24 BLV participants to envision a more accurate and inclusive annotation system. Surprisingly, many participants viewed the high-accuracy technology not just as an annotation system but also as a tool for precise last-few-meters navigation. Guided by participant feedback, we developed NaviNote, which combines vision-based high-precision localization with an agentic architecture to enable voice-based annotation authoring and navigation. Evaluating NaviNote with 18 BLV participants showed that it significantly improved navigation performance and supported users in understanding and annotating their surroundings. Based on these findings, we discuss design considerations for future accessible annotation authoring systems.

cs.HC↗

arXiv2026

ARGaze: Autoregressive Transformers for Online Egocentric Gaze Estimation

Online egocentric gaze estimation predicts where a camera wearer is looking from first-person video using only past and current frames, a task essential for augmented reality and assistive technologies. Unlike third-person gaze estimation, this setting lacks explicit head or eye signals, requiring models to infer current visual attention from sparse, indirect cues such as hand-object interactions and salient scene content. We observe that gaze exhibits strong temporal continuity during goal-directed activities: knowing where a person looked recently provides a powerful prior for predicting where they look next. Inspired by vision-conditioned autoregressive decoding in vision-language models, we propose ARGaze, which reformulates gaze estimation as sequential prediction: at each timestep, a transformer decoder predicts current gaze by conditioning on (i) current visual features and (ii) a fixed-length Gaze Context Window of recent gaze target estimates. This design enforces causality and enables bounded-resource streaming inference. We achieve state-of-the-art performance across multiple egocentric benchmarks under online evaluation, with extensive ablations validating that autoregressive modeling with bounded gaze history is critical for robust prediction. We will release our source code and pre-trained models.

cs.CV↗

arXiv2026

LLM-FSM: Scaling Large Language Models for Finite-State Reasoning in RTL Code Generation

Finite-state reasoning, the ability to understand and implement state-dependent behavior, is central to hardware design. In this paper, we present LLM-FSM, a benchmark that evaluates how well large language models (LLMs) can recover finite-state machine (FSM) behavior from natural-language specifications and translate it into correct register transfer-level (RTL) implementations. Unlike prior specification-to-RTL benchmarks that rely on manually constructed examples, LLM-FSM is built through a fully automated pipeline. LLM-FSM first constructs FSM with configurable state counts and constrained transition structures. It then prompts LLMs to express each FSM in a structured YAML format with an application context, and to further convert that YAML into a natural-language (NL) specification. From the same YAML, our pipeline synthesizes the reference RTL and testbench in a correct-by-construction manner. All 1,000 problems are verified using LLM-based and SAT-solver-based checks, with human review on a subset. Our experiments show that even the strongest LLMs exhibit sharply declining accuracy as FSM complexity increases. We further demonstrate that training-time scaling via supervised fine-tuning (SFT) generalizes effectively to out-of-distribution (OOD) tasks, while increasing test-time compute improves reasoning reliability. Finally, LLM-FSM remains extensible by allowing its FSM complexity to scale with future model capabilities.

cs.AI↗

arXiv2026

RaZeR: Pushing the Limits of NVFP4 Quantization with Redundant Zero Remapping

The recently introduced NVFP4 format demonstrates remarkable performance and memory benefits for quantized large language model (LLM) inference. However, we observe two types of redundancy in NVFP4 encoding: (1) The FP4 element format naturally exposes an unused quantization value due to its sign-magnitude representation that contains both positive and negative zeros. (2) The FP8 block scaling factor has an unused sign bit because it is always positive. Additionally, we find that LLM weights are more tolerant to a lower-precision block scaling factor. Based on these observations, we propose Redundant Zero Remapping (RaZeR), an enhanced numerical format that pushes the limits of NVFP4 for more accurate LLM quantization under the same memory footprint. RaZeR leverages the redundant bits of the block scaling factor to adaptively remap the redundant FP4 zero to additional quantization values with improved accuracy. To demonstrate the practicality of RaZeR, we design efficient GPU kernels for RaZeR-quantized LLM inference and propose novel hardware to natively support this. Extensive experiments validate RaZeR's superior performance for 4-bit LLM quantization. For example, relative to native NVFP4, RaZeR reduces the average perplexity loss by 34.6% and 31.2% under weight-only and weight-activation quantization, respectively. Code is available at: https://github.com/yc2367/NVFP4-RaZeR.

cs.LG↗

arXiv2026

Investigating $Ωϕ$ Interaction and Correlation Functions

In this work, we investigate the interaction between the $Ω$ baryon and the $s\bar{s}$ meson within the framework of the quark delocalization color screening model. The spectra calculations show that no bound state is formed in any of the considered channels, while the scattering indicates that the $Ωϕ$ interaction with $J^{P}=1/2^{-}$ is weakly attractive. As for the $Ωϕ$ interactions with $J^{P}=3/2^{-}$ and $5/2^{-}$, as well as the $Ωη^{\prime}$ interaction with $J^{P}=3/2^{-}$, they are all repulsive. After an investigation on the femtoscopic correlation functions, we find that, due to the spin-averaging effect, the overall $Ωϕ$ correlation function exhibits a weak dependence on the source size, which provides a crucial significance of our model for future experimental examinations in relativistic heavy-ion collisions.

hep-ph↗

arXiv2025

Dynamical study of hidden-strange pentaquarks as analogs of the hidden-charm states

Motivated by the recent BESIII experiment~\cite{BESIII:2024muk} searching for hidden-strange exotic hadrons, we perform a systematic theoretical study of the hidden-strange pentaquark system within the framework of the quark delocalization color screening model (QDCSM) and the resonating group method (RGM). Our results demonstrate that the channel coupling effect plays a decisive role in the formation of bound and resonance states. It not only significantly enhances the short-range attraction but also induces essential attractive contributions from pion exchange. We predict three bound states with masses of $1759$ MeV, $2000$ MeV, and $2407$ MeV. Furthermore, we report the existence of a hidden-strange pentaquark resonance state, $ΞK^{\ast}$, with quantum numbers $I(J^{P})=0(1/2^{-})$. This resonance is identified in the $S$-wave scattering phase shifts of the $Λη_{s}$ and $Λϕ$ open channels, with a predicted mass in the range of $2204\text{--}2208$ MeV. By accounting for both the scattering width from channel coupling and the intrinsic decay width of the constituent $K^{\ast}$, the total decay width is estimated to be $55\text{--}63$ MeV. These theoretical predictions provide important guidance for future experimental searches for such exotic states at facilities like BESIII.

hep-ph↗

arXiv2025

Investigation of Resonances in the $Σ({1/2}^{-})$ System Based on the Chiral Quark Model

In this work, we investigate the resonance structures in the $Σ(1/2^-)$ system from both three-quark and five-quark perspectives within the framework of the chiral quark model. An accurate few-body computational approach, the Gaussian Expansion Method, is employed to construct the orbital wave functions of multiquark states. To reduce the model dependence on parameters, we fit two sets of parameters to check the stability of the results. The calculations show that our results remain stable despite changes in the parameters. In the three-quark calculations, two $Σ(1/2^-)$ states are obtained with energies around 1.8~GeV, which are good candidates for the experimentally observed $Σ(1750)$ and $Σ(1900)$. In the five-quark configuration, several stable resonance states are identified, including $Σπ$, $N \bar{K}$, and $N \bar{K}^{*}$. These resonance states survive the channel-coupling calculations under the complex-scaling framework and manifest as stable structures. Our results support the existence of a two-pole structure for the $Σ(1/2^-)$ system, predominantly composed of $Σπ$ and $N \bar{K}$ configurations, analogous to the well-known $Λ(1380)$-$Λ(1405)$ ($Σπ$-$N \bar{K}$) system. On the other hand, although the energy of the $N \bar{K}^{*}$ configuration is close to that of $Σ(1750)$ and $Σ(1900)$, the obtained width is not consistent with the experimental values. This suggests that the $N \bar{K}^{*}$ state needs to mix with three-quark components to better explain the experimental $Σ(1750)$ and $Σ(1900)$ states. According to our decay width calculations, the predicted two resonance states are primarily composed of $Σπ$ and $N \bar{K}$, with their main decay channel being $Λπ$.

hep-ph↗

arXiv2025

How2Compress: Scalable and Efficient Edge Video Analytics via Adaptive Granular Video Compression

With the rapid proliferation of the Internet of Things, video analytics has become a cornerstone application in wireless multimedia sensor networks. To support such applications under bandwidth constraints, learning-based adaptive quantization for video compression have demonstrated strong potential in reducing bitrate while maintaining analytical accuracy. However, existing frameworks often fail to fully exploit the fine-grained quality control enabled by modern blockbased video codecs, leaving significant compression efficiency untapped. In this paper, we present How2Compress, a simple yet effective framework designed to enhance video compression efficiency through precise, fine-grained quality control at the macroblock level. How2Compress is a plug-and-play module and can be seamlessly integrated into any existing edge video analytics pipelines. We implement How2Compress on the H.264 codec and evaluate its performance across diverse real-world scenarios. Experimental results show that How2Compress achieves up to $50.4\%$ bitrate savings and outperforms baselines by up to $3.01\times$ without compromising accuracy, demonstrating its practical effectiveness and efficiency. Code is available at https://github.com/wyhallenwu/how2compress and a reproducible docker image at https://hub.docker.com/r/wuyuheng/how2compress.

cs.MM↗

arXiv2025

SafeCoop: Unravelling Full Stack Safety in Agentic Collaborative Driving

Collaborative driving systems leverage vehicle-to-everything (V2X) communication across multiple agents to enhance driving safety and efficiency. Traditional V2X systems take raw sensor data, neural features, or perception results as communication media, which face persistent challenges, including high bandwidth demands, semantic loss, and interoperability issues. Recent advances investigate natural language as a promising medium, which can provide semantic richness, decision-level reasoning, and human-machine interoperability at significantly lower bandwidth. Despite great promise, this paradigm shift also introduces new vulnerabilities within language communication, including message loss, hallucinations, semantic manipulation, and adversarial attacks. In this work, we present the first systematic study of full-stack safety and security issues in natural-language-based collaborative driving. Specifically, we develop a comprehensive taxonomy of attack strategies, including connection disruption, relay/replay interference, content spoofing, and multi-connection forgery. To mitigate these risks, we introduce an agentic defense pipeline, which we call SafeCoop, that integrates a semantic firewall, language-perception consistency checks, and multi-source consensus, enabled by an agentic transformation function for cross-frame spatial alignment. We systematically evaluate SafeCoop in closed-loop CARLA simulation across 32 critical scenarios, achieving 69.15% driving score improvement under malicious attacks and up to 67.32% F1 score for malicious detection. This study provides guidance for advancing research on safe, secure, and trustworthy language-driven collaboration in transportation systems. Our project page is https://xiangbogaobarry.github.io/SafeCoop.

cs.CV↗

arXiv2025

On the Role of Temperature Sampling in Test-Time Scaling

Large language models (LLMs) can improve reasoning at inference time through test-time scaling (TTS), where multiple reasoning traces are generated and the best one is selected. Prior work shows that increasing the number of samples K steadily improves accuracy. In this paper, we demonstrate that this trend does not hold indefinitely: at large K, further scaling yields no gains, and certain hard questions remain unsolved regardless of the number of traces. Interestingly, we find that different sampling temperatures solve different subsets of problems, implying that single-temperature scaling explores only part of a model's potential. We therefore propose scaling along the temperature dimension, which enlarges the reasoning boundary of LLMs. Averaged over Qwen3 (0.6B, 1.7B, 4B, 8B) and five representative reasoning benchmarks (AIME 2024/2025, MATH500, LiveCodeBench, Hi-ToM), temperature scaling yields an additional 7.3 points over single-temperature TTS. Temperature scaling also enables base models to reach performance comparable to reinforcement learning (RL)-trained counterparts, without additional post-training. We further provide a comprehensive analysis of this phenomenon and design a multi-temperature voting method that reduces the overhead of temperature scaling. Overall, our findings suggest that TTS is more powerful than previously thought, and that temperature scaling offers a simple and effective way to unlock the latent potential of base models.

cs.AI↗

arXiv2025

DEL-ToM: Inference-Time Scaling for Theory-of-Mind Reasoning via Dynamic Epistemic Logic

Theory-of-Mind (ToM) tasks pose a unique challenge for large language models (LLMs), which often lack the capability for dynamic logical reasoning. In this work, we propose DEL-ToM, a framework that improves verifiable ToM reasoning through inference-time scaling rather than architectural changes. Our approach decomposes ToM tasks into a sequence of belief updates grounded in Dynamic Epistemic Logic (DEL), enabling structured and verifiable dynamic logical reasoning. We use data generated automatically via a DEL simulator to train a verifier, which we call the Process Belief Model (PBM), to score each belief update step. During inference, the PBM evaluates candidate belief traces from the LLM and selects the highest-scoring one. This allows LLMs to allocate extra inference-time compute to yield more transparent reasoning. Experiments across model scales and benchmarks show that DEL-ToM consistently improves performance, demonstrating that verifiable belief supervision significantly enhances LLMs' ToM capabilities without retraining. Code is available at https://github.com/joel-wu/DEL-ToM.

cs.AI↗

arXiv2025

SATBench: Benchmarking LLMs' Logical Reasoning via Automated Puzzle Generation from SAT Formulas

We introduce SATBench, a benchmark for evaluating the logical reasoning capabilities of large language models (LLMs) through logical puzzles derived from Boolean satisfiability (SAT) problems. Unlike prior work that focuses on inference rule-based reasoning, which often involves deducing conclusions from a set of premises, our approach leverages the search-based nature of SAT problems, where the objective is to find a solution that fulfills a specified set of logical constraints. Each instance in SATBench is generated from a SAT formula, then translated into a puzzle using LLMs. The generation process is fully automated and allows for adjustable difficulty by varying the number of clauses. All 2100 puzzles are validated through both LLM-based and solver-based consistency checks, with human validation on a subset. Experimental results show that even the strongest model, o4-mini, achieves only 65.0% accuracy on hard UNSAT problems, close to the random baseline of 50%. Our error analysis reveals systematic failures such as satisfiability bias, context inconsistency, and condition omission, highlighting limitations of current LLMs in search-based logical reasoning. Our code and data are publicly available at https://github.com/Anjiang-Wei/SATBench

cs.AI↗

arXiv2025

A Fully Open and Generalizable Foundation Model for Ultrasound Clinical Applications

Artificial intelligence (AI) that can effectively learn ultrasound representations by integrating multi-source data holds significant promise for advancing clinical care. However, the scarcity of large labeled datasets in real-world clinical environments and the limited generalizability of task-specific models have hindered the development of generalizable clinical AI models for ultrasound applications. In this study, we present EchoCare, a novel ultrasound foundation model for generalist clinical use, developed via self-supervised learning on our curated, publicly available, large-scale dataset EchoCareData. EchoCareData comprises 4.5 million ultrasound images, sourced from over 23 countries across 5 continents and acquired via a diverse range of distinct imaging devices, thus encompassing global cohorts that are multi-center, multi-device, and multi-ethnic. Unlike prior studies that adopt off-the-shelf vision foundation model architectures, we introduce a hierarchical classifier into EchoCare to enable joint learning of pixel-level and representation-level features, capturing both global anatomical contexts and local ultrasound characteristics. With minimal training, EchoCare outperforms state-of-the-art comparison models across 10 representative ultrasound benchmarks of varying diagnostic difficulties, spanning disease diagnosis, lesion segmentation, organ detection, landmark prediction, quantitative regression, imaging enhancement and report generation. The code and pretrained model are publicly released, rendering EchoCare accessible for fine-tuning and local adaptation, supporting extensibility to additional applications. EchoCare provides a fully open and generalizable foundation model to boost the development of AI technologies for diverse clinical ultrasound applications.

cs.CV↗

arXiv2025

CodeARC: Benchmarking Reasoning Capabilities of LLM Agents for Inductive Program Synthesis

Inductive program synthesis, or programming by example, requires synthesizing functions from input-output examples that generalize to unseen inputs. While large language model agents have shown promise in programming tasks guided by natural language, their ability to perform inductive program synthesis is underexplored. Existing evaluation protocols rely on static sets of examples and held-out tests, offering no feedback when synthesized functions are incorrect and failing to reflect real-world scenarios such as reverse engineering. We propose CodeARC, the Code Abstraction and Reasoning Challenge, a new evaluation framework where agents interact with a hidden target function by querying it with new inputs, synthesizing candidate functions, and iteratively refining their solutions using a differential testing oracle. This interactive setting encourages agents to perform function calls and self-correction based on feedback. We construct the first large-scale benchmark for general-purpose inductive program synthesis, featuring 1114 functions. Among 18 models evaluated, o3-mini performs best with a success rate of 52.7%, highlighting the difficulty of this task. Fine-tuning LLaMA-3.1-8B-Instruct on curated synthesis traces yields up to a 31% relative performance gain. CodeARC provides a more realistic and challenging testbed for evaluating LLM-based program synthesis and inductive reasoning. Our code, data, and models are publicly available at https://github.com/Anjiang-Wei/CodeARC

cs.PL↗

arXiv2025

AirV2X: Unified Air-Ground Vehicle-to-Everything Collaboration

While multi-vehicular collaborative driving demonstrates clear advantages over single-vehicle autonomy, traditional infrastructure-based V2X systems remain constrained by substantial deployment costs and the creation of "uncovered danger zones" in rural and suburban areas. We present AirV2X-Perception, a large-scale dataset that leverages Unmanned Aerial Vehicles (UAVs) as a flexible alternative or complement to fixed Road-Side Units (RSUs). Drones offer unique advantages over ground-based perception: complementary bird's-eye-views that reduce occlusions, dynamic positioning capabilities that enable hovering, patrolling, and escorting navigation rules, and significantly lower deployment costs compared to fixed infrastructure. Our dataset comprises 6.73 hours of drone-assisted driving scenarios across urban, suburban, and rural environments with varied weather and lighting conditions. The AirV2X-Perception dataset facilitates the development and standardized evaluation of Vehicle-to-Drone (V2D) algorithms, addressing a critical gap in the rapidly expanding field of aerial-assisted autonomous driving systems. The dataset and development kits are open-sourced at https://github.com/taco-group/AirV2X-Perception.

cs.CV↗

arXiv2025

LangCoop: Collaborative Driving with Language

Multi-agent collaboration holds great promise for enhancing the safety, reliability, and mobility of autonomous driving systems by enabling information sharing among multiple connected agents. However, existing multi-agent communication approaches are hindered by limitations of existing communication media, including high bandwidth demands, agent heterogeneity, and information loss. To address these challenges, we introduce LangCoop, a new paradigm for collaborative autonomous driving that leverages natural language as a compact yet expressive medium for inter-agent communication. LangCoop features two key innovations: Mixture Model Modular Chain-of-thought (M$^3$CoT) for structured zero-shot vision-language reasoning and Natural Language Information Packaging (LangPack) for efficiently packaging information into concise, language-based messages. Through extensive experiments conducted in the CARLA simulations, we demonstrate that LangCoop achieves a remarkable 96\% reduction in communication bandwidth (< 2KB per message) compared to image-based communication, while maintaining competitive driving performance in the closed-loop evaluation. Our project page and code are at https://xiangbogaobarry.github.io/LangCoop/.

cs.RO↗

arXiv2025

Sensitivity Meets Sparsity: The Impact of Extremely Sparse Parameter Patterns on Theory-of-Mind of Large Language Models

This paper investigates the emergence of Theory-of-Mind (ToM) capabilities in large language models (LLMs) from a mechanistic perspective, focusing on the role of extremely sparse parameter patterns. We introduce a novel method to identify ToM-sensitive parameters and reveal that perturbing as little as 0.001% of these parameters significantly degrades ToM performance while also impairing contextual localization and language understanding. To understand this effect, we analyze their interaction with core architectural components of LLMs. Our findings demonstrate that these sensitive parameters are closely linked to the positional encoding module, particularly in models using Rotary Position Embedding (RoPE), where perturbations disrupt dominant-frequency activations critical for contextual processing. Furthermore, we show that perturbing ToM-sensitive parameters affects LLM's attention mechanism by modulating the angle between queries and keys under positional encoding. These insights provide a deeper understanding of how LLMs acquire social reasoning abilities, bridging AI interpretability with cognitive science. Our results have implications for enhancing model alignment, mitigating biases, and improving AI systems designed for human interaction.

cs.CL↗

arXiv2025

Search for doubly charmed dibaryons in baryon-baryon scattering

We perform a systematical investigation of the doubly charmed dibaryon system with quantum numbers $IJ=01$, and strangeness numbers $S=0$, $-2$ and $-4$ in the framework of the chiral quark model. Two resonance states with strangeness numbers $S=-2$ is obtained in the $ΛΩ_{cc}$ scattering channel, which are $Ξ_{cc}^{\ast}Ξ$ with resonance mass 5081 MeV and decay width 0.3 MeV, and the $Ξ_{c}Ξ_{c}^{\ast}$ state with the mass 5213 MeV and decay width 19.8 MeV, respectively. These two predicted charmed dibaryon candidates are worth searching for experimentally. Besides, we would like to emphasize that the multi-channel coupling calculation is important to confirm the existence of multiquark states. The coupling can shift the energy of the resonance, give the width to the resonance and even destroy the resonance. Therefore, to provide the necessary information for experiments to search for exotic hadron states, the coupling calculation between the bound channels and open channels is indispensable.

hep-ph↗