Searcharxiv⌕ Search

arXiv subjects

Zijian Chen

Publications and source records attributed to Zijian Chen.

At least 19 recordsLinked to original sources

Where Do Vision-Language Models Get Left and Right? Position-Code Counterfactuals in Qwen2.5-VL, Qwen3-VL and InternVL3

A vision-language model may take left-right direction from the orientation of image content within patches, its vision transformer's (ViT) position code, or token positions and order in its language model (LLM). Ablations show the ViT code is necessary; we test sufficiency by mirroring only a position code. In Qwen2.5-VL-7B, Qwen3-VL-4B and InternVL3-2B, mirroring ViT width positions reproduces the flipped-image answer on 84.3-93.2% of CV-Bench left-right questions (Qwen2.5-VL-3B: 87.5%, partly a left fallback). With two options, this also counts disruption; a half-grid shift flips most answers too. With the ViT code intact, LLM-side id and token-order mirrors change few left-right answers in all three series (a weak id test in Qwen2.5-VL). In Qwen2.5-VL-7B the effect sits in the global-attention ViT blocks, mostly the last. On What'sUp, flipping content and token order, not the position code, keeps more on/under than left-right answers at the flipped-image answer. Post hoc analyses qualify these results. On CV-Bench items with located objects, random permutations flip 18.9-31.0 points more answers when reversing the objects' order than when keeping it. On What'sUp, in Qwen2.5-VL and Qwen3-VL, the share of the left-right changes made by random permutations that land on the mirror relation is mostly below chance. Counted as returns to the original answer, the What'sUp contrast of support with left-right relations is clear only in InternVL3-2B. Most remaining errors give the same relation whichever object is asked about, suggesting that repair may target relation binding, object localization or use of the reference object.

cs.CV↗

LG-Bench: A Graph-Structured Evaluation Benchmark for Life Sciences

Traditional evaluation benchmarks reduce inherently interconnected scientific knowledge in life sciences into flat lists of questions, disregarding the underlying topological structure of the knowledge. We introduce LG-Bench, the first graph-structured benchmark for life sciences, featuring over 10,000 high-quality multiple-choice questions across medicine, biology, and chemistry. Our approach constructs a weighted evaluation graph using bidirectional matching and semantic similarity algorithms, where nodes represent questions and edge weights capture their semantic relationships. Leveraging this graph topology, we design two novel evaluation metrics. The Global Coherence Score (GCS) measures a model's consistency within semantically related neighborhoods, while Knowledge Balance Score (KBS) analyzes how model errors are distributed across the graph to reveal conceptual blind spots. LG-Bench facilitates fine-grained comparison of LLMs by surfacing differences in conceptual coherence and patterns of knowledge organization across models. Our framework shifts the evaluation paradigm from flat accuracy metrics to structure-aware analysis, offering a new lens for diagnosing and improving LLM performance in the life sciences domain.

cs.AI↗

From Learner Behavior to Reusable Skills for Effective and Efficient Learner Simulation

Learner simulation aims to reproduce how a particular learner behaves on new tasks. Although Large Language Models (LLMs) can generate increasingly fine-grained learning behaviors, existing approaches often need to repeatedly process a growing interaction history to reconstruct the learner. This introduces additional context and inference costs and makes the acquired learner-specific simulation capability difficult to reuse across different LLMs. We therefore propose Learner2Skill, which externalizes the simulation capability acquired from historical interactions into a persistent and reusable Simulation Skill. The Skill captures the learner's current learning state and recurring response patterns, evolves as new real interactions arrive, and can be adapted to a new LLM through lightweight executor calibration without reconstructing the learner from scratch. Experiments show that Learner2Skill more faithfully reproduces fine-grained learner behavior while reducing overall token cost, and that the same constructed Skills can be effectively reused across different LLM executors.

cs.AI↗

From Pixel to Coding: Evaluating the Figure Reproduction Capabilities of MLLMs

Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities in both visual understanding and code generation. However, existing benchmarks typically evaluate these two modalities in isolation, lacking a dedicated assessment of their unification, i.e., how a model can perceive complex visual structures and synthesize them into precise, executable code. Moreover, current visual code generation benchmarks often rely on simplified layouts within single programming environments, falling short of evaluating true unified multimodal reasoning. To bridge this gap, we propose FigCodeBench, a comprehensive framework for rigorously evaluating MLLMs on figure reproduction, integrating multimodal comprehension and generation. We first design a systematic dataset construction pipeline, resulting in a total of 6,194 instances that cover 7 functional categories and 4 types of programming languages. We further categorize figure reproduction into three tiers with visual and code complexity modeling, specifically targeting complex structural reasoning, varying aspect ratios, and dense geometric constraints. We introduce a multi-dimensional evaluation protocol, encompassing visual fidelity and syntactic isomorphism, that aligns highly with the Mean Machine Opinion Score (MMOS) and human preferences. Based on our framework, we conducted extensive experiments on 24 widely used proprietary and open-source MLLMs (e.g., Gemini 3.1 Pro, GPT-5.4, and Kimi-K2.5), where we observed a universal, non-linear performance cliff across different programming languages and difficulty scenarios for all models, and gained several insights, such as the significant metric decline in rigid declarative languages.

cs.CV↗

Invisible in Space, Visible in Time: Motion Vision CAPTCHA against GUI Agents

Most existing visual CAPTCHAs remain spatially solvable: the required information is exposed by static appearance, local structure, and interface state. This assumption is weakened by advances in multimodal large language models (MLLMs) and Graphical User Interface (GUI) agents, which exhibit strong visual perception, reasoning, and browser interaction capabilities. We propose Motion Vision CAPTCHA (MVCAP), a hierarchical motion-based CAPTCHA framework in which target semantics are instantiated as motion-defined foreground structures and become recoverable only through temporal segregation from a dynamically evolving background. Built on this shared principle, MVCAP is instantiated in three perceptually progressive levels: coherent motion, structural motion, and biological motion. To evaluate this framework, we introduce MVCAP-Bench, a browser-based benchmark with 600 live CAPTCHA instances, together with a matched foreground-only control benchmark, MVCAP-Bench-FG. We evaluate humans, Browser Use agents, native computer use agents, and a supplementary offline VQA setting derived from the same instances. Results reveal a substantial human--agent gap: on the full MVCAP-Bench, human accuracy reaches 99.6%, whereas the best GUI agent achieves only 16.8%, close to the six-way chance level. The foreground-only control further shows that the key difficulty comes from dynamic background camouflage rather than answer format or browser interaction alone. These findings identify a measurable human--agent perception gap and position MVCAP-Bench as a benchmark for studying motion-defined perception in current agents.

cs.CV↗

SciMIF: Understanding Multimodal Instruction Following in Scientific Domains

Understanding instruction-following capabilities in scientific domains is essential for effectively leveraging Multimodal Large Language Models (MLLMs) to advance the development of scientific fields. In this work, we introduce SciMIF, a novel benchmark designed to evaluate the capability of MLLMs in following complex scientific instructions. Specifically, based on an extensive analysis of 22 distinct tasks across 5 representative scientific disciplines, we propose a comprehensive taxonomy comprising 10 constraint groups that captures both general functional requirements and discipline-specific characteristics. Guided by this taxonomy, we develop a high-fidelity instruction injection pipeline to systematically augment existing scientific datasets. We conduct comprehensive experiments on multiple state-of-the-art closed-source and open-source MLLMs. Our findings reveal significant performance disparities across different scientific disciplines, with chemistry posing greater challenges for current MLLMs. Furthermore, we observe that increasing the model scale does not yield corresponding improvements in constraint adherence, and current models still struggle severely with fine-grained constraints and instructions requiring the deep application of disciplinary knowledge. SciMIF fills the current void in evaluating multimodal instruction adherence within scientific domains, laying a crucial foundation for future enhancements of MLLMs in rigorous scientific applications. Data and code will be released at https://github.com/shenye7436/SciMIF .

cs.AI↗

Unfolding the Interdisciplinary Complexities of Climate Science: Fuxi-Climate Foundational Model

Climate research and decision-making require integrating evidence across physical processes, socio-economic dynamics and policy responses. Large language models (LLMs) have been explored for accessing and synthesizing climate knowledge, but their ability to support structured interdisciplinary reasoning is still limited. Here we present the Fuxi-Climate Foundation Model (CFM), a climate-specialized LLM designed to support consistent reasoning across domains. CFM maintains more stable analytical behavior as interdisciplinary complexity increases, whereas performance in other models becomes more variable. On expert-designed climate transition tasks, CFM produces more structured analyses that explicitly address trade-offs and uncertainty, achieving 45% trade-off coverage and 47.27% uncertainty-aware reasoning. These results indicate that CFM can support more realistic analysis of climate risks and transition pathways, and provide a basis for agent-based systems to explore complex policy and decision scenarios. The model is openly available at https://huggingface.co/SII-yuning/cfm.

cs.CY↗

Beyond Single Character: Evaluating MLLMs for Sentence-Level Oracle Bone Inscription Understanding

Existing AI-assisted oracle bone inscription (OBI) visual recognition and understanding studies mainly focus on character-level, ignoring the long-form textual coherence and contextual dependencies embedded in complete divination charges. Recently, the powerful visual perception capabilities of multimodal large language models (MLLMs) have opened new possibilities for OBI information processing. In this work, we introduce S-OBI, a novel benchmark for evaluating MLLMs in Sentence-level OBI understanding. Instead of using noisy and incomplete rubbings as the visual input, S-OBI synthesizes clear and standardized sentence-level OBI instances through glyph substitution and composition. According to 95 original rubbings with translations that have been identified, corrected, and verified by experts, we replace characters in the original rubbings with corresponding clean glyph samples sourced from existing OBI datasets while preserving the overall inscriptional structure and semantic organization. This mitigates the influence of low-level distortions and enables a more focused evaluation of sentence-level OBI understanding. Based on this, we design semantic matching, semantic slot extraction, and contextual reasoning tasks and obtain 695 question-answer pairs. Experiments reveal the inferiority of contemporary MLLMs on sentence-level OBI understanding. In particular, visual perception errors in unmasked regions propagate through the reasoning chain, leading to erroneous predictions for masked characters, which indicates that sentence-level OBI understanding in current models remains strongly dependent on character-level recognition. Overall, S-OBI provides a diagnostic benchmark for evaluating whether MLLMs can move beyond isolated character recognition toward structured inscription-level understanding.

cs.CV↗

UniDial-EvalKit: A Unified Toolkit for Evaluating Multi-Faceted Conversational Abilities

Benchmarking large language models (LLMs) and agents in multi-turn interactive scenarios is essential for understanding their practical capabilities. However, existing evaluation protocols are highly heterogeneous, differing significantly in dataset formats, model interfaces, and evaluation pipelines, which severely impedes systematic comparison. In this work, we present UniDial-EvalKit (UDE), a unified evaluation toolkit for assessing interactive AI systems. The core contribution of UDE lies in its holistic unification: it standardizes heterogeneous data formats into a universal schema, streamlines complex evaluation pipelines through a modular architecture, and aligns metric calculations under a hierarchical scoring aggregation. It also supports efficient large-scale evaluation through parallel generation and scoring, as well as checkpoint resume to eliminate redundant computation. Leveraging UDE, we conduct an extensive evaluation across diverse multi-dimensional benchmarks. Our empirical analysis shows that no single system consistently outperforms others across all benchmarks, while current memory agents often fail to surpass full-context baselines. Further analyses highlight several future directions, including benchmark deduplication and more adaptive memory architectures.

cs.CL↗

GeoR-Bench: Evaluating Geoscience Visual Reasoning

Geoscience intelligence is expected to understand, reason about, and predict earth system changes to support human decision-making in critical domains such as disaster response, climate adaptation and environmental protection. Although current research has shown promising progress on specific geoscience tasks, such as remote sensing interpretation, geographic question-answering, existing benchmarks remain largely task-specific which failing to capture the open-ended real world geoscience problems. As a result, it remains unclear how far current AI systems are from achieving genuine geoscience intelligence. To address this gap, we present \textbf{GeoR-Bench}, a \underline{Bench}mark for evaluating \underline{Geo}science visual \underline{R}easoning through reasoning informed visual editing tasks. GeoR-Bench contains 440 curated samples spanning 6 geoscience categories and 24 task types, covering earth observation imagery and structured scientific representations such as maps and diagrams. We evaluate outputs along three dimensions, including reasoning, consistency, and quality. Benchmark results of 21 closed- and open-source multimodal models reveal that geoscience reasoning remains a critical bottleneck. The highest-performing model achieves 42.7\% overall strict accuracy, while the best open-source models only get 10.3\%. Notably, the visual consistency and image quality of the outputs frequently surpass their scientific accuracy. Ultimately, these findings indicate that current models generate superficially plausible results but fail to capture underlying earth science processes.

cs.CV↗

Sulcal Pattern Matching with the Wasserstein Distance

We present the unified computational framework for modeling the sulcal patterns of human brain obtained from the magnetic resonance images. The Wasserstein distance is used to align the sulcal patterns nonlinearly. These patterns are topologically different across subjects making the pattern matching a challenge. We work out the mathematical details and develop the gradient descent algorithms for estimating the deformation field. We further quantify the image registration performance. This method is applied in identifying the differences between male and female sulcal patterns.

q-bio.NC↗

SafeSci: Safety Evaluation of Large Language Models in Science Domains and Beyond

The success of large language models (LLMs) in scientific domains has heightened safety concerns, prompting numerous benchmarks to evaluate their scientific safety. Existing benchmarks often suffer from limited risk coverage and a reliance on subjective evaluation. To address these problems, we introduce SafeSci, a comprehensive framework for safety evaluation and enhancement in scientific contexts. SafeSci comprises SafeSciBench, a multi-disciplinary benchmark with 0.25M samples, and SafeSciTrain, a large-scale dataset containing 1.5M samples for safety enhancement. SafeSciBench distinguishes between safety knowledge and risk to cover extensive scopes and employs objective metrics such as deterministically answerable questions to mitigate evaluation bias. We evaluate 24 advanced LLMs, revealing critical vulnerabilities in current models. We also observe that LLMs exhibit varying degrees of excessive refusal behaviors on safety-related issues. For safety enhancement, we demonstrate that fine-tuning on SafeSciTrain significantly enhances the safety alignment of models. Finally, we argue that knowledge is a double-edged sword, and determining the safety of a scientific question should depend on specific context, rather than universally categorizing it as safe or unsafe. Our work provides both a diagnostic tool and a practical resource for building safer scientific AI systems.

cs.LG↗

ORBIT: Scalable and Verifiable Data Generation for Search Agents on a Tight Budget

Search agents, which integrate language models (LMs) with web search, are becoming crucial for answering complex user queries. Constructing training datasets for deep research tasks, involving multi-step retrieval and reasoning, remains challenging due to expensive human annotation, or cumbersome prerequisites. In this work, we introduce ORBIT, a training dataset with 20K reasoning-intensive queries with short verifiable answers, generated using a frugal framework without relying on paid API services. The modular framework relies on four stages: seed creation, question-answer pair generation, and two stages of verification: self and external. ORBIT spans 15 domains and each training pair requires 4-5 reasoning steps, with external search verification required from the complete web. We train Qwen3-4B as the base model on ORBIT using GRPO and evaluate it on Wikipedia question answering tasks. Extensive experiment results demonstrate that ORBIT-4B achieves strong performance among sub-4B LLMs as search agents, proving the utility of synthetic datasets. Our framework, code and datasets are open-sourced and available publicly.

cs.CL↗

A$^3$: Towards Advertising Aesthetic Assessment

Advertising images significantly impact commercial conversion rates and brand equity, yet current evaluation methods rely on subjective judgments, lacking scalability, standardized criteria, and interpretability. To address these challenges, we present A^3 (Advertising Aesthetic Assessment), a comprehensive framework encompassing four components: a paradigm (A^3-Law), a dataset (A^3-Dataset), a multimodal large language model (A^3-Align), and a benchmark (A^3-Bench). Central to A^3 is a theory-driven paradigm, A^3-Law, comprising three hierarchical stages: (1) Perceptual Attention, evaluating perceptual image signals for their ability to attract attention; (2) Formal Interest, assessing formal composition of image color and spatial layout in evoking interest; and (3) Desire Impact, measuring desire evocation from images and their persuasive impact. Building on A^3-Law, we construct A^3-Dataset with 120K instruction-response pairs from 30K advertising images, each richly annotated with multi-dimensional labels and Chain-of-Thought (CoT) rationales. We further develop A^3-Align, trained under A^3-Law with CoT-guided learning on A^3-Dataset. Extensive experiments on A^3-Bench demonstrate that A^3-Align achieves superior alignment with A^3-Law compared to existing models, and this alignment generalizes well to quality advertisement selection and prescriptive advertisement critique, indicating its potential for broader deployment. Dataset, code, and models can be found at: https://github.com/euleryuan/A3-Align.

cs.CV↗

Benchmarking Cross-Scale Perception Ability of Large Multimodal Models in Material Science

Unraveling the hierarchical structure-property relationships is the central challenge of materials science, necessitating the interpretation of data across vast physical scales from micro to macro. Despite the rapid integration of Large Multimodal Models (LMMs) into scientific workflows, existing scientific benchmarks primarily focus on general chart interpretation or isolated common-sense reasoning, failing to capture reasoning ability across intricate physical dimensions. To address this, we introduce CSMBench, a dataset comprising 1,041 high-quality figures curated from premier journals up to September 2025. CSMBench categorizes data into four scientifically distinct regimes: atomic, micro, meso, and macro scales, strictly aligning with the focus and definitions in materials study. Through open-ended figure description and multiple-choice caption matching tasks, we evaluate state-of-the-art open-source and closed-source models. Our analysis identifies that performance varies significantly across physical scales due to the distinct visual characteristics, highlighting the limitations of current generalist models and identifying critical directions for achieving hierarchical and accurate understanding in materials research. The CSMBench is publicly released at: https://huggingface.co/datasets/lututu/CSMBench.

cs.DL↗

InterveneBench: Benchmarking LLMs for Intervention Reasoning and Causal Study Design in Real Social Systems

Causal inference in social science relies on end-to-end, intervention-centered research-design reasoning grounded in real-world policy interventions, but current benchmarks fail to evaluate this capability of large language models (LLMs). We present InterveneBench, a benchmark designed to assess such reasoning in realistic social settings. Each instance in InterveneBench is derived from an empirical social science study and requires models to reason about policy interventions and identification assumptions without access to predefined causal graphs or structural equations. InterveneBench comprises 744 peer-reviewed studies across diverse policy domains. Experimental results show that state-of-the-art LLMs struggle under this setting. To address this limitation, we further propose a multi-agent framework, STRIDES. It achieves significant performance improvements over state-of-the-art reasoning models. Our code and data are available at https://github.com/Sii-yuning/STRIDES.

cs.CY↗

AgentIR: Reasoning-Aware Retrieval for Deep Research Agents

Deep Research agents are rapidly emerging as primary consumers of modern retrieval systems. Unlike human users who issue and refine queries without documenting their intermediate thought processes, Deep Research agents generate explicit natural language reasoning before each search call, revealing rich intent and contextual information that existing retrievers entirely ignore. To exploit this overlooked signal, we introduce: (1) Reasoning-Aware Retrieval, a retrieval paradigm that jointly embeds the agent's reasoning trace alongside its query; and (2) DR-Synth, a data synthesis method that generates Deep Research retriever training data from standard QA datasets. We demonstrate that both components are independently effective, and their combination yields a trained embedding model, AgentIR-4B, with substantial gains. On the challenging BrowseComp-Plus benchmark, AgentIR-4B achieves 68\% accuracy with the open-weight agent Tongyi-DeepResearch, compared to 50\% with conventional embedding models twice its size, and 37\% with BM25. Code and data are available at: https://texttron.github.io/AgentIR/.

cs.CL↗

STAR : Bridging Statistical and Agentic Reasoning for Large Model Performance Prediction

As comprehensive large model evaluation becomes prohibitively expensive, predicting model performance from limited observations has become essential. However, existing statistical methods struggle with pattern shifts, data sparsity, and lack of explanation, while pure LLM methods remain unreliable. We propose STAR, a framework that bridges data-driven STatistical expectations with knowledge-driven Agentic Reasoning. STAR leverages specialized retrievers to gather external knowledge and embeds semantic features into Constrained Probabilistic Matrix Factorization (CPMF) to generate statistical expectations with uncertainty. A reasoning module guided by Expectation Violation Theory (EVT) then refines predictions through intra-family analysis, cross-model comparison, and credibility-aware aggregation, producing adjustments with traceable explanations. Extensive experiments show that STAR consistently outperforms all baselines on both score-based and rank-based metrics, delivering a 14.46% gain in total score over the strongest statistical method under extreme sparsity, with only 1--2 observed scores per test model.

cs.AI↗