arXiv subjects

Jin Cui

Publications and source records attributed to Jin Cui.

At least 19 recordsLinked to original sources

arXiv2026

HAFI-VLM: A Frequency Perspective for Diagnosing and Enhancing Visual Perception in Vision-Language Models

Vision-language models (VLMs) remain unreliable when predictions require fine-grained visual evidence. We identify a previously overlooked cause: spectral response rigidity. Despite substantial frequency variation across images and tasks, pretrained vision encoders exhibit persistent, encoder-specific layerwise spectral profiles that change only marginally under downstream fine-tuning. Since pretrained vision encoders only receive images, they cannot adapt spectral extraction to the evidence required by the current query. We therefore propose HAFI-VLM, which introduces a task-conditioned frequency pathway while preserving the pretrained semantic representation. Hierarchical Adaptive Frequency Injection (HAFI) retrieves complementary low-, mid-, and high-frequency evidence at multiple encoder depths using text-modulated, spatially aligned cross-attention. A Visual Enrichment Layer Adapter further recalibrates shallow LLM attention to effectively utilize the enriched visual tokens. Experiments on LLaVA-1.5 and Qwen2.5-VL demonstrate consistent improvements in general VQA, text-rich understanding, and hallucination robustness, outperforming representation-level enhancement methods and most resolution- or cropping-based approaches without additional high-resolution encoding. Mechanistic analyses show that HAFI restores task-dependent spectral allocation while retaining semantic attention, establishing frequency enrichment as a distinct and effective route for improving VLM perception.

cs.CV↗

arXiv2026

Look Where It Matters: Adaptive Visual Refinement for Vision-Language-Action Models

Visual representations of VLA models remain unreliable for spatially precise robotic manipulation. We uncover that vision encoders in VLAs also exhibit attention artifacts previously documented in generic Vision Transformers, and further show that, in embodied policies, these artifacts are closely associated with spatial perception capabilities acquired during post-training. As the encoder learns task-relevant information such as object location, depth ordering, and local geometry, limited global-token capacity causes part of this information to spill into low-information patch tokens. We introduce AtVLA, a framework that inserts learnable register tokens into the visual encoder. Trained end-to-end using only embodied data and the original action objective, these registers emerge as dedicated carriers of embodied spatial information, while the remaining patch tokens recover clean and spatially faithful attention distributions crucial for precise target localization and fine-grained contact. Clean attention restores reliable localization, but cannot recover geometric details lost in low-resolution observations. AtVLA therefore couples attention rectification with uncertainty-gated local refinement. The action expert samples multiple action chunks and estimates uncertainty from their disagreement; only for uncertain predictions, action-conditioned attention rollout identifies the task-relevant region, which is cropped, re-encoded at high resolution, and appended to the cached prefix for refined action generation. Across LIBERO, SimplerEnv, and a challenging single-view real-world benchmark, AtVLA improves the average LIBERO success rate from 94.2% to 98.4% and real-world success from 46.5% to 69.0%. The cropping is triggered on approximately 30% of replanning steps, resulting in only 1.4-1.6x the total computation of the base model under the representative deployment setting.

cs.RO↗

arXiv2026

GLOBE: Trajectory-Aligned Gradient Matching with Structured SparseOptimization for Coreset Selection

On-device training of deep neural networks is fundamentally constrained by the computational and memory costs of large-scale datasets. Coreset selection offers a practical solution by retaining only a compact subset of real training samples. However, existing gradient-based methods commonly rely on gradients computed at a single model snapshot and employ greedy or pursuit-based selection procedures, limiting their ability to capture evolving optimization dynamics and handle strongly correlated samples. We propose GLOBE (Gradient Local-Balanced Extraction), a trajectory-aligned coreset selection framework that formulates sample selection as a globally optimized sparse weighting problem. GLOBE represents each sample by a gradient trajectory constructed across multiple training checkpoints, thereby capturing its influence throughout different stages of optimization. To preserve the training behavior of the full dataset, we introduce a multi-order matching objective that jointly aligns the first-order mean and projected uncentered second-order moments of gradient trajectories. GLOBE further combines Group LASSO, Elastic Net regularization, and nonnegative budget constraints to induce group- and sample-level sparsity while stabilizing the weights of correlated trajectories. Finally, class-balanced Top-K selection maintains adequate category coverage under limited sampling budgets. Experiments across six benchmarks and five evaluation architectures demonstrate that GLOBE consistently outperforms existing coreset selection methods in downstream test accuracy, particularly at low retention ratios. These results highlight the effectiveness of combining dynamic gradient information, multi-order distribution matching, and structured sparsity for data-efficient learning.

cs.LG↗

arXiv2026

Distill What RGB Can Recover: Privileged 3D Evidence for RGB-Only Vision-Language Models

3D scene understanding requires reasoning about entity existence, spatial layout, and object relations, yet RGB images alone often provide insufficient 3D cues. Existing 3D-VLMs commonly rely on depth or 3D-position-aware inputs at inference time, introducing additional acquisition, reconstruction, or annotation costs that limit RGB-only deployment. We therefore study how training-time 3D evidence can be converted into spatial reasoning capabilities retained under RGB-only inference. We propose a privileged-evidence distillation framework that constructs a distillable teacher through a unified evidence interface and controlled residual injection, and transfers its knowledge to a deployable student receiving only RGB images and questions through logit and structured representation distillation. To avoid imitating teacher signals unsupported by RGB, we further introduce evidence-sensitivity-guided distillation, which uses corrupted evidence to identify highly evidence-dependent targets and down-weight their supervision. We also define a recoverability decomposition based on the matched baseline, teacher, and student, separating privileged gains into RGB-recoverable improvements and residual teacher advantages. Across four benchmarks, the teacher achieves the best result on 7 of 11 reported metrics among the compared methods. The RGB-only student outperforms its matched baseline on all 11 metrics, including gains of 10.4 ScanQA CIDEr and 19.1 Scan2Cap CIDEr@0.5, without additional inference-time inputs. These results validate the effectiveness of training-time privileged 3D evidence distillation for both teacher performance and deployable RGB-only spatial reasoning. Separately, our matched baseline-teacher-student analysis characterizes privileged-gain transfer across evidence types and spatial skills.

cs.CV↗

arXiv2026

ECHO: Continuous Hierarchical Memory for Vision-Language-Action Models

Memory capacity is a critical factor determining the performance of Vision-Language-Action (VLA) models in long-horizon manipulation tasks. Existing memory-augmented architectures primarily rely on linear or flat storage, lacking structural priors for manipulation categories and hierarchical organization. This deficiency hinders efficient experience retrieval and limits generalization to unseen long-horizon task compositions. Inspired by the hierarchical organization of human experience, we propose ECHO (Experience Consolidation and Hierarchical Organization), a novel memory framework operating within a Continuous Hierarchical Space. By employing a hyperbolic autoencoder, ECHO maps VLA hidden states into this space. Leveraging hyperbolic metrics and entailment constraint mechanisms, experience vectors are organized into a semantic memory tree that supports efficient top-down retrieval. In parallel, a background consolidation mechanism continuously refines the memory tree through geometric interpolation and structural splitting, supporting virtual memory synthesis in the continuous space. We integrate ECHO into the $\pi_0$ foundation model. Evaluations on LIBERO and preliminary real-world experiments demonstrate the effectiveness of our approach, notably achieving a 12.8% absolute improvement in execution success rate over the $\pi_0$ baseline on LIBERO-Long, while improving compositional generalization on cross-suite unseen long-horizon tasks.

cs.RO↗

arXiv2026

Retrieve, Integrate, and Synthesize: Spatial-Semantic Grounded Latent Visual Reasoning

Multimodal Large Language Models (MLLMs) have made remarkable progress on vision-language reasoning, yet most methods still compress visual evidence into discrete textual thoughts, creating an information bottleneck for fine-grained perception. Recent latent visual reasoning methods attempt to reason in continuous hidden states, but we find that they suffer from insufficient manifold compatibility: latent trajectories drift away from pretrained reasoning circuits, collapse into instance-agnostic patterns, and are often bypassed during answer generation. To address these issues, we propose RIS (Retrieve, Integrate, and Synthesize), a spatial-semantic grounded framework that develops latent reasoning as a compatible extension of pretrained MLLM computation. We first construct a step-wise grounded reasoning dataset with bounding boxes and region-specific semantic descriptions. Built on this supervision, RIS anchors latent tokens to both spatial and semantic evidence, enforces their causal role through a progressive attention bottleneck, and introduces short language transition tokens to bridge synthesized latent states back to vocabulary-aligned decoding. Experiments on V*, HRBench4K, HRBench8K, MMVP, and BLINK show consistent improvements over closed/open-source and latent reasoning baselines. Further analyses demonstrate that RIS learns diverse, interpretable, and progressively integrated latent trajectories, offering a practical path toward faithful internal visual reasoning in MLLMs.

cs.CL↗

arXiv2026

"The Whole Is Greater Than the Sum of Its Parts": A Compatibility-Aware Multi-Teacher CoT Distillation Framework

Chain-of-Thought (CoT) reasoning empowers Large Language Models (LLMs) with remarkable capabilities but typically requires prohibitive parameter scales. CoT distillation has emerged as a promising paradigm to transfer reasoning prowess into compact Student Models (SLMs), but existing approaches often rely on a solitary teacher, capping the student's potential since individual LLMs often exhibit distinct capability biases and may suffer from catastrophic forgetting. While leveraging diverse teachers seems appealing, effectively fusing their supervisions remains challenging: teacher-student incompatibility risks amplifying hallucinations, and passive supervision fails to ensure genuine logic internalization. To address this, we introduce COMPACT, a framework that adaptively fuses supervisions from different teachers by dynamically weighting teacher gradients based on the student's real-time compatibility evaluated by a multi-dimensional metric: (1) Graph-based Consensus to filter misleading rationales by identifying mainstream reasoning paths; (2) Mutual-Information-based Adaptability to detect "epiphany moments" for genuinely understanding the reasoning process rather than merely imitating; and (3) Loss-based Difficulty to assess student receptivity to the teacher's guidance and prevent negative transfer. Extensive experiments and latent space analysis demonstrate that COMPACT effectively integrates diverse reasoning capabilities without damaging the model's original knowledge structure, achieving state-of-the-art performance on various benchmarks while mitigating catastrophic forgetting.

cs.CL↗

arXiv2026

MIND: From Passive Mimicry to Active Reasoning through Capability-Aware Multi-Perspective CoT Distillation

While Large Language Models (LLMs) have emerged with remarkable capabilities in complex tasks through Chain-of-Thought reasoning, practical resource constraints have sparked interest in transferring these abilities to smaller models. However, achieving both domain performance and cross-domain generalization remains challenging. Existing approaches typically restrict students to following a single golden rationale and treat different reasoning paths independently. Due to distinct inductive biases and intrinsic preferences, alongside the student's evolving capacity and reasoning preferences during training, a teacher's "optimal" rationale could act as out-of-distribution noise. This misalignment leads to a degeneration of the student's latent reasoning distribution, causing suboptimal performance. To bridge this gap, we propose MIND, a capability-adaptive framework that transitions distillation from passive mimicry to active cognitive construction. We synthesize diverse teacher perspectives through a novel "Teaching Assistant" network. By employing a Feedback-Driven Inertia Calibration mechanism, this network utilizes inertia-filtered training loss to align supervision with the student's current adaptability, effectively enhancing performance while mitigating catastrophic forgetting. Extensive experiments demonstrate that MIND achieves state-of-the-art performance on both in-distribution and out-of-distribution benchmarks, and our sophisticated latent space analysis further confirms the mechanism of reasoning ability internalization.

cs.CL↗

arXiv2025

FAST: Topology-Aware Frequency-Domain Distribution Matching for Coreset Selection

Coreset selection compresses large datasets into compact, representative subsets, reducing the energy and computational burden of training deep neural networks. Existing methods are either: (i) DNN-based, which are tied to model-specific parameters and introduce architectural bias; or (ii) DNN-free, which rely on heuristics lacking theoretical guarantees. Neither approach explicitly constrains distributional equivalence, largely because continuous distribution matching is considered inapplicable to discrete sampling. Moreover, prevalent metrics (e.g., MSE, KL, CE, MMD) cannot accurately capture higher-order moment discrepancies, leading to suboptimal coresets. In this work, we propose FAST, the first DNN-free distribution-matching coreset selection framework that formulates the coreset selection task as a graph-constrained optimization problem grounded in spectral graph theory and employs the Characteristic Function Distance (CFD) to capture full distributional information in the frequency domain. We further discover that naive CFD suffers from a "vanishing phase gradient" issue in medium and high-frequency regions; to address this, we introduce an Attenuated Phase-Decoupled CFD. Furthermore, for better convergence, we design a Progressive Discrepancy-Aware Sampling strategy that progressively schedules frequency selection from low to high, preserving global structure before refining local details and enabling accurate matching with fewer frequencies while avoiding overfitting. Extensive experiments demonstrate that FAST significantly outperforms state-of-the-art coreset selection methods across all evaluated benchmarks, achieving an average accuracy gain of 9.12%. Compared to other baseline coreset methods, it reduces power consumption by 96.57% and achieves a 2.2x average speedup, underscoring its high performance and energy efficiency.

stat.ML↗

arXiv2025

Accurate and Interpretable Postmenstrual Age Prediction via Multimodal Large Language Model

Accurate estimation of postmenstrual age (PMA) at scan is crucial for assessing neonatal development and health. While deep learning models have achieved high accuracy in predicting PMA from brain MRI, they often function as black boxes, offering limited transparency and interpretability in clinical decision support. In this work, we address the dual challenge of accuracy and interpretability by adapting a multimodal large language model (MLLM) to perform both precise PMA prediction and clinically relevant explanation generation. We introduce a parameter-efficient fine-tuning (PEFT) strategy using instruction tuning and Low-Rank Adaptation (LoRA) applied to the Qwen2.5-VL-7B model. The model is trained on four 2D cortical surface projection maps derived from neonatal MRI scans. By employing distinct prompts for training and inference, our approach enables the MLLM to handle a regression task during training and generate clinically relevant explanations during inference. The fine-tuned model achieves a low prediction error with a 95 percent confidence interval of 0.78 to 1.52 weeks, while producing interpretable outputs grounded in developmental features, marking a significant step toward transparent and trustworthy AI systems in perinatal neuroscience.

cs.AI↗

arXiv2025

RecGPT Technical Report

Recommender systems are among the most impactful applications of artificial intelligence, serving as critical infrastructure connecting users, merchants, and platforms. However, most current industrial systems remain heavily reliant on historical co-occurrence patterns and log-fitting objectives, i.e., optimizing for past user interactions without explicitly modeling user intent. This log-fitting approach often leads to overfitting to narrow historical preferences, failing to capture users' evolving and latent interests. As a result, it reinforces filter bubbles and long-tail phenomena, ultimately harming user experience and threatening the sustainability of the whole recommendation ecosystem. To address these challenges, we rethink the overall design paradigm of recommender systems and propose RecGPT, a next-generation framework that places user intent at the center of the recommendation pipeline. By integrating large language models (LLMs) into key stages of user interest mining, item retrieval, and explanation generation, RecGPT transforms log-fitting recommendation into an intent-centric process. To effectively align general-purpose LLMs to the above domain-specific recommendation tasks at scale, RecGPT incorporates a multi-stage training paradigm, which integrates reasoning-enhanced pre-alignment and self-training evolution, guided by a Human-LLM cooperative judge system. Currently, RecGPT has been fully deployed on the Taobao App. Online experiments demonstrate that RecGPT achieves consistent performance gains across stakeholders: users benefit from increased content diversity and satisfaction, merchants and the platform gain greater exposure and conversions. These comprehensive improvement results across all stakeholders validates that LLM-driven, intent-centric design can foster a more sustainable and mutually beneficial recommendation ecosystem.

cs.IR↗

arXiv2025

Analyzing Patient Daily Movement Behavior Dynamics Using Two-Stage Encoding Model

In the analysis of remote healthcare monitoring data, time series representation learning offers substantial value in uncovering deeper patterns of patient behavior, especially given the fine temporal granularity of the data. In this study, we focus on a dataset of home activity records from people living with Dementia. We propose a two-stage self-supervised learning approach. The first stage involves converting time-series activities into text strings, which are then encoded by a fine-tuned language model. In the second stage, these time-series vectors are bi-dimensionalized for applying PageRank method, to analyze latent state transitions to quantitatively assess participants behavioral patterns and identify activity biases. These insights, combined with diagnostic data, aim to support personalized care interventions.

cs.AI↗

arXiv2025

Using MRNet to Predict Lunar Rock Categories Detected by Chang'e 5 Probe

China's Chang'e 5 mission has been a remarkable success, with the chang'e 5 lander traveling on the Oceanus Procellarum to collect images of the lunar surface. Over the past half century, people have brought back some lunar rock samples, but its quantity does not meet the need for research. Under current circumstances, people still mainly rely on the analysis of rocks on the lunar surface through the detection of lunar rover. The Oceanus Procellarum, chosen by Chang'e 5 mission, contains various kind of rock species. Therefore, we first applied to the National Astronomical Observatories of the China under the Chinese Academy of Sciences for the Navigation and Terrain Camera (NaTeCam) of the lunar surface image, and established a lunar surface rock image data set CE5ROCK. The data set contains 100 images, which randomly divided into training, validation and test set. Experimental results show that the identification accuracy testing on convolutional neural network (CNN) models like AlexNet or MobileNet is about to 40.0%. In order to make full use of the global information in Moon images, this paper proposes the MRNet (MoonRockNet) network architecture. The encoding structure of the network uses VGG16 for feature extraction, and the decoding part adds dilated convolution and commonly used U-Net structure on the original VGG16 decoding structure, which is more conducive to identify more refined but more sparsely distributed types of lunar rocks. We have conducted extensive experiments on the established CE5ROCK data set, and the experimental results show that MRNet can achieve more accurate rock type identification, and outperform other existing mainstream algorithms in the identification performance.

cs.CV↗

arXiv2025

Two-Stage Representation Learning for Analyzing Movement Behavior Dynamics in People Living with Dementia

In remote healthcare monitoring, time series representation learning reveals critical patient behavior patterns from high-frequency data. This study analyzes home activity data from individuals living with dementia by proposing a two-stage, self-supervised learning approach tailored to uncover low-rank structures. The first stage converts time-series activities into text sequences encoded by a pre-trained language model, providing a rich, high-dimensional latent state space using a PageRank-based method. This PageRank vector captures latent state transitions, effectively compressing complex behaviour data into a succinct form that enhances interpretability. This low-rank representation not only enhances model interpretability but also facilitates clustering and transition analysis, revealing key behavioral patterns correlated with clinicalmetrics such as MMSE and ADAS-COG scores. Our findings demonstrate the framework's potential in supporting cognitive status prediction, personalized care interventions, and large-scale health monitoring.

cs.LG↗

arXiv2024

Enhancing High-order Interaction Awareness in LLM-based Recommender Model

Large language models (LLMs) have demonstrated prominent reasoning capabilities in recommendation tasks by transforming them into text-generation tasks. However, existing approaches either disregard or ineffectively model the user-item high-order interactions. To this end, this paper presents an enhanced LLM-based recommender (ELMRec). We enhance whole-word embeddings to substantially enhance LLMs' interpretation of graph-constructed interactions for recommendations, without requiring graph pre-training. This finding may inspire endeavors to incorporate rich knowledge graphs into LLM-based recommenders via whole-word embedding. We also found that LLMs often recommend items based on users' earlier interactions rather than recent ones, and present a reranking solution. Our ELMRec outperforms state-of-the-art (SOTA) methods in both direct and sequential recommendations.

cs.IR↗

arXiv2024

RDRec: Rationale Distillation for LLM-based Recommendation

Large language model (LLM)-based recommender models that bridge users and items through textual prompts for effective semantic reasoning have gained considerable attention. However, few methods consider the underlying rationales behind interactions, such as user preferences and item attributes, limiting the reasoning capability of LLMs for recommendations. This paper proposes a rationale distillation recommender (RDRec), a compact model designed to learn rationales generated by a larger language model (LM). By leveraging rationales from reviews related to users and items, RDRec remarkably specifies their profiles for recommendations. Experiments show that RDRec achieves state-of-the-art (SOTA) performance in both top-N and sequential recommendations. Our source code is released at https://github.com/WangXFng/RDRec.

cs.CL↗

arXiv2024

CaDRec: Contextualized and Debiased Recommender Model

Recommender models aimed at mining users' behavioral patterns have raised great attention as one of the essential applications in daily life. Recent work on graph neural networks (GNNs) or debiasing methods has attained remarkable gains. However, they still suffer from (1) over-smoothing node embeddings caused by recursive convolutions with GNNs, and (2) the skewed distribution of interactions due to popularity and user-individual biases. This paper proposes a contextualized and debiased recommender model (CaDRec). To overcome the over-smoothing issue, we explore a novel hypergraph convolution operator that can select effective neighbors during convolution by introducing both structural context and sequential context. To tackle the skewed distribution, we propose two strategies for disentangling interactions: (1) modeling individual biases to learn unbiased item embeddings, and (2) incorporating item popularity with positional encoding. Moreover, we mathematically show that the imbalance of the gradients to update item embeddings exacerbates the popularity bias, thus adopting regularization and weighting schemes as solutions. Extensive experiments on four datasets demonstrate the superiority of the CaDRec against state-of-the-art (SOTA) methods. Our source code and data are released at https://github.com/WangXFng/CaDRec.

cs.IR↗

arXiv2024

NFARec: A Negative Feedback-Aware Recommender Model

Graph neural network (GNN)-based models have been extensively studied for recommendations, as they can extract high-order collaborative signals accurately which is required for high-quality recommender systems. However, they neglect the valuable information gained through negative feedback in two aspects: (1) different users might hold opposite feedback on the same item, which hampers optimal information propagation in GNNs, and (2) even when an item vastly deviates from users' preferences, they might still choose it and provide a negative rating. In this paper, we propose a negative feedback-aware recommender model (NFARec) that maximizes the leverage of negative feedback. To transfer information to multi-hop neighbors along an optimal path effectively, NFARec adopts a feedback-aware correlation that guides hypergraph convolutions (HGCs) to learn users' structural representations. Moreover, NFARec incorporates an auxiliary task - predicting the feedback sentiment polarity (i.e., positive or negative) of the next interaction - based on the Transformer Hawkes Process. The task is beneficial for understanding users by learning the sentiment expressed in their previous sequential feedback patterns and predicting future interactions. Extensive experiments demonstrate that NFARec outperforms competitive baselines. Our source code and data are released at https://github.com/WangXFng/NFARec.

cs.IR↗