arXiv subjects

Bin Hu

Publications and source records attributed to Bin Hu.

At least 73 records · Page 4Linked to original sources

arXiv2026

First, Do No Harm: AI Supervisor Scaffolds Novice Growth in Counselor Education

The most dangerous mistakes a novice counselor makes are not the obvious ones: they are utterances that sound caring while quietly violating professional ethics and leaving vulnerable clients less protected. We build an AI supervisor that does not replace novice counselors, but grows them-teaching them to internalize ethical violations they would otherwise never notice. What makes this supervisor non-trivial is not detection but teaching: it must locate the ethical-violating utterance, diagnose the ethical violation against APA principles, and deliver feedback that explains not just what went wrong, but why it is risky and how to respond differently. The core obstacle is that (1) ethical violations are by nature unlabeled in real clinical data, and (2) existing AI counselors trained only to match correct answers will never learn to teach. We resolve both at once: a controllable AI novice that intentionally enacts predefined mistake categories makes supervision labels a natural byproduct of generation, yielding ETHICSCAFF, a 9,915-instance human-in-the-loop dataset; and GRPO under a Novice Growth Reward (NGR) optimizes the supervisor not for answer correctness but for whether a weaker novice model actually improves after reading its explanation. Experiments show that a novice guided by our supervisor outperforms an unguided peer on clinical metrics, and that teaching-oriented optimization via NGR further sharpens the supervisor's own ethical detection. In a user study with novice counseling-psychology students, participants show significant self-efficacy gains across all eight assessed competencies after receiving AI supervisory feedback, demonstrating that the scaffold transfers from simulation to real-world practice.

cs.CL↗

arXiv2026

TSRBench: A Comprehensive Multi-task Multi-modal Time Series Reasoning Benchmark for Generalist Models

Time series are ubiquitous in real-world scenarios and crucial for applications ranging from energy management to traffic control. Consequently, the ability to reason over time series is a fundamental skill for generalist models to solve complex problems. However, current benchmarks for generalist models largely overlook this dimension. To bridge this gap, we introduce TSRBench, a comprehensive multi-modal benchmark designed to stress-test the full spectrum of time series reasoning capabilities. TSRBench features: i) a diverse set of 4125 problems from 14 domains, and is categorized into 4 major dimensions: Perception, Reasoning, Prediction, and Decision-Making. ii) 15 tasks from the 4 dimensions evaluating essential reasoning capabilities (e.g., numerical reasoning). Through extensive experiments, we evaluate over 30 leading proprietary and open-source LLMs, VLMs, and TSLLMs within TSRBench. Our findings reveal that: i) scaling laws hold for perception and reasoning but break down for prediction; ii) strong reasoning does not guarantee accurate context-aware forecasting, indicating a decoupling between semantic understanding and numerical prediction; and iii) despite the complementary nature of textual and visual forms of time series as inputs, current multimodal models fail to effectively fuse them for reciprocal performance gains. TSRBench provides a standardized evaluation platform that not only highlights existing challenges but also offers valuable insights to advance generalist models. Our code and dataset are available at https://tsrbench.github.io/.

cs.AI↗

arXiv2026

On the Implicit Reward Overfitting and the Low-rank Dynamics in RLVR

Recent extensive research has demonstrated that the enhanced reasoning capabilities acquired by models through Reinforcement Learning with Verifiable Rewards (RLVR) are primarily concentrated within the rank-1 components. Predicated on this observation, we employed Periodic Rank-1 Substitution and identified a counterintuitive phenomenon: RLVR may exhibit implicit reward overfitting to the training dataset. Specifically, the model can achieve satisfactory performance on the test set even when its rewards remain relatively low during the training process. Furthermore, we characterize three distinct properties of RL training: (1) The effective rank-1 component in RLVR don't maintain other model knowledge except mathematical reasoning capability. (2) RLVR fundamentally functions by optimizing a specific singular spectrum. The distribution of singular values of almost all linear layers in RLVR-trained model behaves like heavy-tailed distribution. (3) the left singular vectors associated with rank-1 components demonstrate a stronger alignment tendency during training, which echoes the discovery that RLVR is optimizing sampling efficiency in essence. Taken together, our findings and analysis further reveal how RLVR shapes model parameters and offer potential insights for improving existing RL paradigms or other training paradigms to implement continual learning.

cs.LG↗

arXiv2026

XiYOLO: Energy-Aware Object Detection via Iterative Architecture Search and Scaling

Object detection on heterogeneous edge devices must satisfy strict energy, latency, and memory constraints while still providing reliable perception for downstream autonomy. Existing energy-aware NAS methods often target limited deployment settings, while real energy remains difficult to optimize because it is highly device-dependent and costly to measure. We address these challenges with an energy-adaptive framework that combines an energy-aware XiResOFA search space, a two-stage energy estimator, and iterative search to identify a single energy-efficient base architecture. We then apply compound scaling to transform this base design into the XiYOLO family across deployment budgets, enabling interpretable accuracy-energy tradeoffs under sparse hardware measurements. Experiments on PascalVOC, COCO, and real-device deployment show that XiYOLO achieves a stronger energy-accuracy tradeoff than YOLO baselines. On PascalVOC, the medium XiYOLO model reaches 86.15 mAP50 while reducing energy relative to YOLOv12m by 20.6% on GPU and 35.9% on NPU. On COCO, XiYOLO reduces energy relative to YOLOv12 by up to 53.7% on GPU and 51.6% on NPU at the small scale. The proposed two-stage estimator also improves sample efficiency over a joint predictor under few-shot adaptation with only 2-20 target-device samples.

cs.CV↗

arXiv2026

Target-depth sensing with metasurface-encoder integrated optoelectronic neural network

Accurate and real-time sensing of targets in three-dimensional (3D) environments is essential for modern machine vision, underpinning emerging technologies such as autonomous systems, robotic manipulation, augmented reality, and intelligent surveillance. However, state-of-the-art 3D sensing approaches typically rely on complex postprocessing of multi-view images or LiDAR point clouds, resulting in considerable computational load, power consumption, and latency. To address these challenges, we propose a metasurface-encoder integrated optoelectronic neural network architecture that compresses 3D information into two-dimensional images by encoding depth using double-helix point spread function generated by a metasurface. The depth-encoded images are captured with a conventional monocular camera and subsequently processed by a lightweight shadow ResNet neural network. We experimentally validate the proposed architecture on the MNIST and Vehicle-Image datasets, achieving high accuracy simultaneously in target classification and depth estimation, thereby enabling real-time target tracking. The framework is readily extendable to other depth- or angle-encoding metasurfaces for multidimensional compression and detection. Our results demonstrate the effectiveness of the meta-optic-encoder/electronic-decoder paradigm in significantly reducing network complexity and computational burden while maintaining strong performance for smart vision sensory applications.

physics.optics↗

arXiv2026

EvoSpark: Endogenous Interactive Agent Societies for Unified Long-Horizon Narrative Evolution

Realizing endogenous narrative evolution in LLM-based multi-agent systems is hindered by the inherent stochasticity of generative emergence. In particular, long-horizon simulations suffer from social memory stacking, where conflicting relational states accumulate without resolution, and narrative-spatial dissonance, where spatial logic detaches from the evolving plot. To bridge this gap, we propose EvoSpark, a framework specifically designed to sustain logically coherent long-horizon narratives within Endogenous Interactive Agent Societies. To ensure consistency, the Stratified Narrative Memory employs a Role Socio-Evolutionary Base as living cognition, dynamically metabolizing experiences to resolve historical conflicts. Complementarily, Generative Mise-en-Scène mechanism enforces Role-Location-Plot alignment, synchronizing character presence with the narrative flow. Underpinning these is the Unified Narrative Operation Engine, which integrates an Emergent Character Grounding Protocol to transform stochastic sparking into persistent characters. This engine establishes a substrate that expands a minimal premise into an open-ended, evolving story world. Experiments demonstrate that EvoSpark significantly outperforms baselines across diverse paradigms, enabling the sustained generation of expressive and coherent narrative experiences.

cs.CL↗

arXiv2026

Glide-in-Place: Foot-Steered Differential-Drive for Hands-Free VR Locomotion

Seated VR locomotion in constrained environments, including homes, offices, and transit settings, calls for hardware that is lightweight and deployable, steering that remains continuous enough for curved motion, and a control channel that leaves the hands free for concurrent interaction. Inspired by the steering logic of self-balancing scooters, we present Glide-in-Place, a seated foot locomotion system that maps per-foot fore-aft pressure to a differential-drive model: the two feet act as virtual wheels whose relative drive continuously determines translation and yaw. This lets users move forward, rotate in place, and follow arcs in one unified vocabulary without hand-held input or discrete mode switches. We evaluated Glide-in-Place in a counterbalanced within-subject study with 16 participants against two baselines: joystick control and a seated walking-in-place technique with discrete snap motions. Across two steering-heavy navigation tasks, zig-zag path following with multitasking and curved-path traversal, Glide-in-Place was consistently faster than Seated-WIP, reduced physical demand, and lowered fatigue-related discomfort without significantly differing from joystick control on total VRSQ. We position Glide-in-Place as a deployable hardware-control design point for constrained seated VR: thin insole sensing, continuous foot steering, and lightweight calibration packaged in one compact artifact.

cs.HC↗

arXiv2026

CCD-CBT: Multi-Agent Therapeutic Interaction for CBT Guided by Cognitive Conceptualization Diagram

Large language models show potential for scalable mental-health support by simulating Cognitive Behavioral Therapy (CBT) counselors. However, existing methods often rely on static cognitive profiles and omniscient single-agent simulation, failing to capture the dynamic, information-asymmetric nature of real therapy. We introduce CCD-CBT, a multi-agent framework that shifts CBT simulation along two axes: 1) from a static to a dynamically reconstructed Cognitive Conceptualization Diagram (CCD), updated by a dedicated Control Agent, and 2) from omniscient to information-asymmetric interaction, where the Therapist Agent must reason from inferred client states. We release CCDCHAT, a synthetic multi-turn CBT dataset generated under this framework. Evaluations with clinical scales and expert therapists show that models fine-tuned on CCDCHAT outperform strong baselines in both counseling fidelity and positive-affect enhancement, with ablations confirming the necessity of dynamic CCD guidance and asymmetric agent design. Our work offers a new paradigm for building theory-grounded, clinically-plausible conversational agents.

cs.CL↗

arXiv2026

Detecting Malicious Intents in Smart Contracts with Pre-trained Programming Language Models

Malicious developer intents in smart contracts constitute significant security threats to decentralized applications, leading to substantial economic losses. Prior work introduced SmartIntentNN, a deep learning model for detecting unsafe developer intents. By combining the Universal Sentence Encoder, a K-means clustering-based intent highlighting mechanism, and a Bidirectional Long Short-Term Memory (BiLSTM) network, the model achieved an F1 score of 0.8633 on an evaluation set of 10,000 real-world smart contracts across ten distinct intent categories. This paper presents SmartIntentV2 (Smart Contract Intent Neural Network Version 2). The primary enhancement is the integration of a BERT-based pre-trained programming language model, which we domain-adaptively pre-train on a dataset of 16,000 real-world smart contracts using a Masked Language Modeling objective. SmartIntentV2 retains the BiLSTM-based multi-label classification network for intent detection. On the same evaluation set of 10,000 smart contracts, it achieves superior performance with an accuracy of 0.9789, precision of 0.9090, recall of 0.9476, and an F1 score of 0.9279, substantially outperforming its predecessor and other baseline models. Notably, SmartIntentV2 also delivers a 65.5% relative improvement in F1 score over GPT-4.1 on this specialized task. These results establish SmartIntentV2 as a new state-of-the-art model for smart contract intent detection.

cs.SE↗

arXiv2026

Time delay measurements with Broken Power Law model

One challenge in strong gravitational lensing cosmography is the measurement of time delays between multiple lensed images, which are essential for constraining the Hubble constant ($H_0$). In this study, we investigate how assumptions about the lens mass profile affect time-delay measurements in lensing systems. Specifically, we implement a Broken Power Law (BPL) mass model within the \textsc{Lenstronomy} framework (Birrer & Amara 2018), which introduces additional flexibility in the radial mass distribution and can phenomenologically capture deviations from a single power-law profile. This model is combined with a numerical approach to compute time delays at the image positions. We validate the BPL implementation using simulated lens systems and compare the results with those from the elliptical power-law (EPL) model. We then apply both model families to the quadruply imaged quasar WGD~2038--4008. Both models provide good fits to the imaging and kinematic data, with a slight preference for the BPL model. When the internal mass-sheet factor is allowed to vary, the inferred Hubble constant in a flat $Λ$CDM cosmology with fixed $Ω_{\rm m}=0.3$ is $H_0 = 75.3^{+23.1}_{-16.3} \ \mathrm{km \ s^{-1} \ Mpc^{-1}}$ for the BPL model and $H_0 = 61.1^{+19.2}_{-13.2} \ \mathrm{km \ s^{-1} \ Mpc^{-1}}$ for the EPL model. For comparison, in the diagnostic case with the internal mass-sheet factor fixed to unity under the same setup, we obtain $H_0 = 74.2^{+20.3}_{-13.8} \ \mathrm{km \ s^{-1} \ Mpc^{-1}}$ for the BPL model and $H_0 = 66.1^{+18.8}_{-12.8} \ \mathrm{km \ s^{-1} \ Mpc^{-1}}$ for the EPL model. This highlights how time-delay cosmography remains sensitive to assumptions about the lens mass profile. With current precision, this difference does not favor one cosmological scenario over another, but underscores the importance of flexible mass modeling.

astro-ph.CO↗

arXiv2026

$\mathcal{H}$-EFTCAMB: A Cobaya-Integrated, Python-Wrapped Extension of EFTCAMB for Covariant Horndeski Gravity

We present $\mathcal{H}\mathtt{-EFTCAMB}$, the official successor to $\mathtt{EFTCAMB}$. The original $\mathtt{EFTCAMB}$ is designed as a consistent and numerically stable implementation of the effective field theory (EFT) of dark energy in the Einstein-Boltzmann code $\mathtt{CAMB}$. On top of this, $\mathcal{H}\mathtt{-EFTCAMB}$ introduces a new Horndeski module that supports computing cosmology for an arbitrary input covariant Horndeski Lagragian. $\mathcal{H}\mathtt{-EFTCAMB}$ supports both mapping the Horndeski theory to an EFT lagrangian to solve in the EFT framework as well as directly solving for the scalar field equations of motion derived from the covariant Lagrangian. The latter approach also works for the cases when the Horndeski field experiences turn-overs, e.g. oscillation, where the EFT approach breaks down. The Horndeski module has been validated by comparing internally with existing models in the original $\mathtt{EFTCAMB}$ and externally with $\mathtt{hi\_class}$. $\mathcal{H}\mathtt{-EFTCAMB}$ features a flexible Python wrapper that is seamlessly integrated into the widely utilized cosmological sampler $\mathtt{Cobaya}$. \heft~is publicly available and serves as a comprehensive tool for testing gravity against the precision data from current and next-generation surveys.

gr-qc↗

arXiv2026

SynPO: Synergizing Descriptiveness and Preference Optimization for Video Detailed Captioning

Fine-grained video captioning aims to generate detailed, temporally coherent descriptions of video content. However, existing methods struggle to capture subtle video dynamics and rich detailed information. In this paper, we leverage preference learning to enhance the performance of vision-language models in fine-grained video captioning, while mitigating several limitations inherent to direct preference optimization (DPO). First, we propose a pipeline for constructing preference pairs that leverages the intrinsic properties of VLMs along with partial assistance from large language models, achieving an optimal balance between cost and data quality. Second, we propose Synergistic Preference Optimization (SynPO), a novel optimization method offering significant advantages over DPO and its variants. SynPO prevents negative preferences from dominating the optimization, explicitly preserves the model's language capability to avoid deviation of the optimization objective, and improves training efficiency by eliminating the need for the reference model. We extensively evaluate SynPO not only on video captioning benchmarks (e.g., VDC, VDD, VATEX) but also across well-established NLP tasks, including general language understanding and preference evaluation, using diverse pretrained models. Results demonstrate that SynPO consistently outperforms DPO variants while achieving 20\% improvement in training efficiency. Code is available at https://github.com/longmalongma/SynPO

cs.AI↗

arXiv2026

Identity-Consistent Video Generation under Large Facial-Angle Variations

Single-view reference-to-video methods often struggle to preserve identity consistency under large facial-angle variations. This limitation naturally motivates the incorporation of multi-view facial references. However, simply introducing additional reference images exacerbates the \textit{copy-paste} problem, particularly the \textbf{\textit{view-dependent copy-paste}} artifact, which reduces facial motion naturalness. Although cross-paired data can alleviate this issue, collecting such data is costly. To balance the consistency and naturalness, we propose $\mathrm{Mv}^2\mathrm{ID}$, a multi-view conditioned framework under in-paired supervision. We introduce a region-masking training strategy to prevent shortcut learning and extract essential identity features by encouraging the model to aggregate complementary identity cues across views. In addition, we design a reference decoupled-RoPE mechanism that assigns distinct positional encoding to video and conditioning tokens for better modeling of their heterogeneous properties. Furthermore, we construct a large-scale dataset with diverse facial-angle variations and propose dedicated evaluation metrics for identity consistency and motion naturalness. Extensive experiments demonstrate that our method significantly improves identity consistency while maintaining motion naturalness, outperforming existing approaches trained with cross-paired data.

cs.CV↗

arXiv2026

Integral Quadratic Constraints for Repeated ReLU

This paper presents a new dynamic integral quadratic constraint (IQC) for the repeated Rectified Linear Unit (ReLU). These dynamic IQCs can be used to analyze stability and induced $\ell_2$-gain performance of discrete-time, recurrent neural networks (RNNs) with ReLU activation functions. These analysis conditions can be incorporated into learning-based controller synthesis methods, which currently rely on static IQCs. We show that our proposed dynamic IQCs for repeated ReLU form a superset of the dynamic IQCs for repeated, slope-restricted nonlinearities. We also prove that the $\ell_2$-gain bounds are nonincreasing with respect to the horizon used in the dynamic IQC filter. A numerical example using a simple (academic) RNN shows that our proposed IQCs lead to less conservative bounds than existing IQCs.

eess.SY↗

arXiv2026

ELASTIC: Efficient Once For All Iterative Search for Object Detection on Microcontrollers

Deploying high-performance object detectors on TinyML platforms poses significant challenges due to tight hardware constraints and the modular complexity of modern detection pipelines. Neural Architecture Search (NAS) offers a path toward automation, but existing methods either restrict optimization to individual modules, sacrificing cross-module synergy, or require global searches that are computationally intractable. We propose ELASTIC (Efficient Once for AlL IterAtive Search for ObjecT DetectIon on MiCrocontrollers), a unified, hardware-aware NAS framework that alternates optimization across modules (e.g., backbone, neck, and head) in a cyclic fashion. ELASTIC introduces a novel Population Passthrough mechanism in evolutionary search that retains high-quality candidates between search stages, yielding faster convergence, up to an 8% final mAP gain, and eliminates search instability observed without population passthrough. In a controlled comparison, empirical results show ELASTIC achieves +4.75% higher mAP and 2x faster convergence than progressive NAS strategies on SVHN, and delivers a +9.09% mAP improvement on PascalVOC given the same search budget. ELASTIC achieves 72.3% mAP on PascalVOC, outperforming MCUNET by 20.9% and TinyissimoYOLO by 16.3%. When deployed on MAX78000/MAX78002 microcontrollers, ELASTICderived models outperform Analog Devices' TinySSD baselines, reducing energy by up to 71.6%, lowering latency by up to 2.4x, and improving mAP by up to 6.99 percentage points across multiple datasets. The experimental videos and codes are available on the project website (https://nail-uh.github.io/elastic.github.io/).

cs.CV↗

arXiv2026

Formation-Aware Adaptive Conformalized Perception for Safe Leader-Follower Multi-Robot Systems

This paper considers the perception safety problem in distributed vision-based leader-follower formations, where each robot uses onboard perception to estimate relative states, track desired setpoints, and keep the leader within its camera field of view (FOV). Safety is challenging due to heteroscedastic perception errors and the coupling between formation maneuvers and visibility constraints. We propose a distributed, formation-aware adaptive conformal prediction method based on Risk-Aware Mondrian CP to produce formation-conditioned uncertainty quantiles. The resulting bounds tighten in high-risk configurations (near FOV limits) and relax in safer regions. We integrate these bounds into a Formation-Aware Conformal CBF-QP with a smooth margin to enforce visibility while maintaining feasibility and tracking performance. Gazebo simulations show improved formation success rates and tracking accuracy over non-adaptive (global) CP baselines that ignore formation-dependent visibility risk, while preserving finite-sample probabilistic safety guarantees. The experimental videos are available on the \href{https://nail-uh.github.io/iros2026.github.io/}{project website}\footnote{Project Website: https://nail-uh.github.io/iros2026.github.io/}.

cs.RO↗

arXiv2026

Doxing via the Lens: Revealing Location-related Privacy Leakage on Multi-modal Large Reasoning Models

Recent advances in multi-modal large reasoning models (MLRMs) have shown significant ability to interpret complex visual content. While these models enable impressive reasoning capabilities, they also introduce novel and underexplored privacy risks. In this paper, we identify a novel category of privacy leakage in MLRMs: Adversaries can infer sensitive geolocation information, such as a user's home address or neighborhood, from user-generated images, including selfies captured in private settings. To formalize and evaluate these risks, we propose a three-level visual privacy risk framework that categorizes image content based on contextual sensitivity and potential for location inference. We further introduce DoxBench, a curated dataset of 500 real-world images reflecting diverse privacy scenarios. Our evaluation across 11 advanced MLRMs and MLLMs demonstrates that these models consistently outperform non-expert humans in geolocation inference and can effectively leak location-related private information. This significantly lowers the barrier for adversaries to obtain users' sensitive geolocation information. We further analyze and identify two primary factors contributing to this vulnerability: (1) MLRMs exhibit strong reasoning capabilities by leveraging visual clues in combination with their internal world knowledge; and (2) MLRMs frequently rely on privacy-related visual clues for inference without any built-in mechanisms to suppress or avoid such usage. To better understand and demonstrate real-world attack feasibility, we propose GeoMiner, a collaborative attack framework that decomposes the prediction process into two stages: clue extraction and reasoning to improve geolocation performance while introducing a novel attack perspective. Our findings highlight the urgent need to reassess inference-time privacy risks in MLRMs to better protect users' sensitive information.

cs.CR↗

arXiv2026

PEGAsus: 3D Personalization of Geometry and Appearance

We present PEGAsus, a new framework capable of generating Personalized 3D shapes by learning shape concepts at both Geometry and Appearance levels. First, we formulate 3D shape personalization as extracting reusable, category-agnostic geometric and appearance attributes from reference shapes, and composing these attributes with text to generate novel shapes. Second, we design a progressive optimization strategy to learn shape concepts at both the geometry and appearance levels, decoupling the shape concept learning process. Third, we extend our approach to region-wise concept learning, enabling flexible concept extraction, with context-aware and context-free losses. Extensive experimental results show that PEGAsus is able to effectively extract attributes from a wide range of reference shapes and then flexibly compose these concepts with text to synthesize new shapes. This enables fine-grained control over shape generation and supports the creation of diverse, personalized results, even in challenging cross-category scenarios. Both quantitative and qualitative experiments demonstrate that our approach outperforms existing state-of-the-art solutions.

cs.CV↗