arXiv subjects

Zihan Gao

Publications and source records attributed to Zihan Gao.

11 recordsLinked to original sources

arXiv2026

Tears or Cheers? Benchmarking LLMs via Culturally Elicited Distinct Affective Responses

Culture serves as a fundamental determinant of human affective processing and profoundly shapes how individuals perceive and interpret emotional stimuli. Despite this intrinsic link extant evaluations regarding cultural alignment within Large Language Models primarily prioritize declarative knowledge such as geographical facts or established societal customs. These benchmarks remain insufficient to capture the subjective interpretative variance inherent to diverse sociocultural lenses. To address this limitation, we introduce CEDAR, a multimodal benchmark constructed entirely from scenarios capturing Culturally \underline{\textsc{E}}licited \underline{\textsc{D}}istinct \underline{\textsc{A}}ffective \underline{\textsc{R}}esponses. To construct CEDAR, we implement a novel pipeline that leverages LLM-generated provisional labels to isolate instances yielding cross-cultural emotional distinctions, and subsequently derives reliable ground-truth annotations through rigorous human evaluation. The resulting benchmark comprises 10,962 instances across seven languages and 14 fine-grained emotion categories, with each language including 400 multimodal and 1,166 text-only samples. Comprehensive evaluations of 17 representative multilingual models reveal a dissociation between language consistency and cultural alignment, demonstrating that culturally grounded affective understanding remains a significant challenge for current models.

cs.CL↗

arXiv2026

$M^2$-VLA: Boosting Vision-Language Models for Generalizable Manipulation via Layer Mixture and Meta-Skills

Current Vision-Language-Action (VLA) models predominantly rely on end-to-end fine-tuning. While effective, this paradigm compromises the inherent generalization capabilities of Vision-Language Models (VLMs) and incurs catastrophic forgetting. To address these limitations, we propose $M^2$-VLA, which demonstrates that a generalized VLM is able to serve as a powerful backbone for robotic manipulation directly. However, it remains a key challenge to bridge the gap between the high-level semantic understanding of VLMs and the precise requirements of robotic control. To overcome this, we introduce the Mixture of Layers (MoL) strategy that selectively extracts task-critical information from dense semantic features. Furthermore, to facilitate efficient trajectory learning under constrained model capacity, we propose a Meta Skill Module (MSM) that integrates strong inductive biases. Extensive experiments in both simulated and real-world environments demonstrate the effectiveness of our approach. Furthermore, generalization and ablation studies validate the architecture's zero-shot capabilities and confirm the contribution of each key component. Our code and pre-trained models will be made publicly available.

cs.RO↗

arXiv2025

Collective Narrative Grounding: Community-Coordinated Data Contributions to Improve Local AI Systems

Large language model (LLM) question-answering systems often fail on community-specific queries, creating "knowledge blind spots" that marginalize local voices and reinforce epistemic injustice. We present Collective Narrative Grounding, a participatory protocol that transforms community stories into structured narrative units and integrates them into AI systems under community governance. Learning from three participatory mapping workshops with N=24 community members, we designed elicitation methods and a schema that retain narrative richness while enabling entity, time, and place extraction, validation, and provenance control. To scope the problem, we audit a county-level benchmark of 14,782 local information QA pairs, where factual gaps, cultural misunderstandings, geographic confusions, and temporal misalignments account for 76.7% of errors. On a participatory QA set derived from our workshops, a state-of-the-art LLM answered fewer than 21% of questions correctly without added context, underscoring the need for local grounding. The missing facts often appear in the collected narratives, suggesting a direct path to closing the dominant error modes for narrative items. Beyond the protocol and pilot, we articulate key design tensions, such as representation and power, governance and control, and privacy and consent, providing concrete requirements for retrieval-first, provenance-visible, locally governed QA systems. Together, our taxonomy, protocol, and participatory evaluation offer a rigorous foundation for building community-grounded AI that better answers local questions.

cs.CL↗

arXiv2025

RoboWheel: A Data Engine from Real-World Human Demonstrations for Cross-Embodiment Robotic Learning

We introduce Robowheel, a data engine that converts human hand object interaction (HOI) videos into training-ready supervision for cross morphology robotic learning. From monocular RGB or RGB-D inputs, we perform high precision HOI reconstruction and enforce physical plausibility via a reinforcement learning (RL) optimizer that refines hand object relative poses under contact and penetration constraints. The reconstructed, contact rich trajectories are then retargeted to cross-embodiments, robot arms with simple end effectors, dexterous hands, and humanoids, yielding executable actions and rollouts. To scale coverage, we build a simulation-augmented framework on Isaac Sim with diverse domain randomization (embodiments, trajectories, object retrieval, background textures, hand motion mirroring), which enriches the distributions of trajectories and observations while preserving spatial relationships and physical plausibility. The entire data pipeline forms an end to end pipeline from video,reconstruction,retargeting,augmentation data acquisition. We validate the data on mainstream vision language action (VLA) and imitation learning architectures, demonstrating that trajectories produced by our pipeline are as stable as those from teleoperation and yield comparable continual performance gains. To our knowledge, this provides the first quantitative evidence that HOI modalities can serve as effective supervision for robotic learning. Compared with teleoperation, Robowheel is lightweight, a single monocular RGB(D) camera is sufficient to extract a universal, embodiment agnostic motion representation that could be flexibly retargeted across embodiments. We further assemble a large scale multimodal dataset combining multi-camera captures, monocular videos, and public HOI corpora for training and evaluating embodied models.

cs.RO↗

arXiv2025

LocalBench: Benchmarking LLMs on County-Level Local Knowledge and Reasoning

Large language models (LLMs) have been widely evaluated on macro-scale geographic tasks, such as global factual recall, event summarization, and regional reasoning. Yet, their ability to handle hyper-local knowledge remains poorly understood. This gap is increasingly consequential as real-world applications, from civic platforms to community journalism, demand AI systems that can reason about neighborhood-specific dynamics, cultural narratives, and local governance. Existing benchmarks fall short in capturing this complexity, often relying on coarse-grained data or isolated references. We present LocalBench, the first benchmark designed to systematically evaluate LLMs on county-level local knowledge across the United States. Grounded in the Localness Conceptual Framework, LocalBench includes 14,782 validated question-answer pairs across 526 U.S. counties in 49 states, integrating diverse sources such as Census statistics, local subreddit discourse, and regional news. It spans physical, cognitive, and relational dimensions of locality. Using LocalBench, we evaluate 13 state-of-the-art LLMs under both closed-book and web-augmented settings. Our findings reveal critical limitations: even the best-performing models reach only 56.8% accuracy on narrative-style questions and perform below 15.5% on numerical reasoning. Moreover, larger model size and web augmentation do not guarantee better performance, for example, search improves Gemini's accuracy by +13.6%, but reduces GPT-series performance by -11.4%. These results underscore the urgent need for language models that can support equitable, place-aware AI systems: capable of engaging with the diverse, fine-grained realities of local communities across geographic and cultural contexts.

cs.CL↗

arXiv2025

A Turing Test for ''Localness'': Conceptualizing, Defining, and Recognizing Localness in People and Machines

As digital platforms increasingly mediate interactions tied to place, ensuring genuine local participation is essential for maintaining trust and credibility in location-based services, community-driven platforms, and civic engagement systems. However, localness is a social and relational identity shaped by knowledge, participation, and community recognition. Drawing on the German philosopher Heidegger's concept of dwelling -- which extends beyond physical presence to encompass meaningful connection to place -- we investigate how people conceptualize and evaluate localness in both human and artificial agents. Using a chat-based interaction paradigm inspired by Turing's Imitation Game and Von Ahn's Games With A Purpose, we engaged 230 participants in conversations designed to examine the cues people rely on to assess local presence. Our findings reveal a multi-dimensional framework of localness, highlighting differences in how locals and nonlocals emphasize various aspects of local identity. We show that people are significantly more accurate in recognizing locals than nonlocals, suggesting that localness is an affirmative status requiring active demonstration rather than merely the absence of nonlocal traits. Additionally, we identify conditions under which artificial agents are perceived as local and analyze participants' sensemaking strategies in evaluating localness. Through predictive modeling, we determine key factors that drive accurate localness judgments. By bridging theoretical perspectives on human-place relationships with practical challenges in digital environments, our work informs the design of location-based services that foster meaningful local engagement. Our findings contribute to a broader understanding of localness as a dynamic and relational construct, reinforcing the importance of dwelling as a process of belonging, recognition, and engagement with place.

cs.HC↗

arXiv2024

Fast and Efficient: Mask Neural Fields for 3D Scene Segmentation

Understanding 3D scenes is a crucial challenge in computer vision research with applications spanning multiple domains. Recent advancements in distilling 2D vision-language foundation models into neural fields, like NeRF and 3DGS, enable open-vocabulary segmentation of 3D scenes from 2D multi-view images without the need for precise 3D annotations. However, while effective, these methods typically rely on the per-pixel distillation of high-dimensional CLIP features, introducing ambiguity and necessitating complex regularization strategies, which adds inefficiency during training. This paper presents MaskField, which enables efficient 3D open-vocabulary segmentation with neural fields from a novel perspective. Unlike previous methods, MaskField decomposes the distillation of mask and semantic features from foundation models by formulating a mask feature field and queries. MaskField overcomes ambiguous object boundaries by naturally introducing SAM segmented object shapes without extra regularization during training. By circumventing the direct handling of dense high-dimensional CLIP features during training, MaskField is particularly compatible with explicit scene representations like 3DGS. Our extensive experiments show that MaskField not only surpasses prior state-of-the-art methods but also achieves remarkably fast convergence. We hope that MaskField will inspire further exploration into how neural fields can be trained to comprehend 3D scenes from 2D models.

cs.CV↗

arXiv2024

Exploring the MBTI distribution among Chinese undergraduate physics students: the influence of family income on career trajectories

This study investigated the distribution of MBTI personality types among physics undergraduates at Zhejiang University and analyzed their career aspirations, family income and mental health status. A comprehensive survey of 68 questions, including 52 multiple-choice questions and 16 short-answer questions, assessed various personality traits and their correlation with academic intent and emotional state. The results showed that INTJ and INTP personality types showed the most significant tendency to pursue academic research. They usually come from middle and above-class backgrounds. Notably, these two personality types accounted for about 41\% of the total surveyed population, while NT types combined accounted for about 58\%. This indicates that although NT personality is more suitable for academic research, students with introverted tendencies are more suitable. Research has further shown that people with ISTJ personalities often exhibit unexpectedly strong academic interests. In addition, individuals who aspire to enter academia generally maintain a stable state of mind, as evidenced by their consistent work efficiency in the face of personal challenges. Interestingly, although some aspiring academics come from financially stable families, most have encountered financial constraints. These results contribute to a deeper understanding of how personality traits and socioeconomic factors influence the academic career paths of physics students.

physics.ed-ph↗

arXiv2024

Development and Characterization of a Novel BaTiO3-Based Material for Medium Temperature Applications

Positive temperature coefficient (PTC) materials are extensively utilized in self-regulating temperature applications. Nonetheless, their applicability is typically constrained to low-temperature ranges, rendering them ineffective in medium temperature environments. This study presents a methodology for the fabrication of an innovative PTC material operational at approximately 353~°C, with a thorough investigation of its Curie temperature and resistivity properties. The material formulation incorporates 4~wt\% carbon black (CB), 0.5~wt\% NBT, and 5~wt\% DOP into a BaTiO$_3$-based matrix. The empirical findings reveal that this material exhibits a notably high PTC strength of 5.8 and a comparatively low resistivity of 590~$Ω\cdot$cm at room temperature. Furthermore, the material demonstrated excellent repeatability in PTC strength after thirty cycles of heating and cooling near the Curie temperature. Consequently, this PTC material is deemed highly effective for applications in cold environments, notably for the preheating and initiation of aircraft engines and auxiliary power units (APUs).

cond-mat.mtrl-sci↗

arXiv2024

Multiplane Prior Guided Few-Shot Aerial Scene Rendering

Neural Radiance Fields (NeRF) have been successfully applied in various aerial scenes, yet they face challenges with sparse views due to limited supervision. The acquisition of dense aerial views is often prohibitive, as unmanned aerial vehicles (UAVs) may encounter constraints in perspective range and energy constraints. In this work, we introduce Multiplane Prior guided NeRF (MPNeRF), a novel approach tailored for few-shot aerial scene rendering-marking a pioneering effort in this domain. Our key insight is that the intrinsic geometric regularities specific to aerial imagery could be leveraged to enhance NeRF in sparse aerial scenes. By investigating NeRF's and Multiplane Image (MPI)'s behavior, we propose to guide the training process of NeRF with a Multiplane Prior. The proposed Multiplane Prior draws upon MPI's benefits and incorporates advanced image comprehension through a SwinV2 Transformer, pre-trained via SimMIM. Our extensive experiments demonstrate that MPNeRF outperforms existing state-of-the-art methods applied in non-aerial contexts, by tripling the performance in SSIM and LPIPS even with three views available. We hope our work offers insights into the development of NeRF-based applications in aerial scenes with limited data.

cs.CV↗

arXiv2022

AIM 2022 Challenge on Instagram Filter Removal: Methods and Results

This paper introduces the methods and the results of AIM 2022 challenge on Instagram Filter Removal. Social media filters transform the images by consecutive non-linear operations, and the feature maps of the original content may be interpolated into a different domain. This reduces the overall performance of the recent deep learning strategies. The main goal of this challenge is to produce realistic and visually plausible images where the impact of the filters applied is mitigated while preserving the content. The proposed solutions are ranked in terms of the PSNR value with respect to the original images. There are two prior studies on this task as the baseline, and a total of 9 teams have competed in the final phase of the challenge. The comparison of qualitative results of the proposed solutions and the benchmark for the challenge are presented in this report.

cs.CV↗