The advancement of vision-language models (VLMs) is hampered by a fragmented landscape of inconsistent and contaminated public datasets. We introduce FineVision, a meticulously collected, curated, and unified corpus of 24 million samples - the largest open resource of its kind. We unify more than 200 sources into 185 subsets via a semi-automated, human-in-the-loop pipeline: automation performs bulk ingestion and schema mapping, while reviewers audit mappings and spot-check outputs to verify faithful consumption of annotations, appropriate formatting and diversity, and safety; issues trigger targeted fixes and re-runs. The workflow further applies rigorous de-duplication within and across sources and decontamination against 66 public benchmarks. FineVision also encompasses agentic/GUI tasks with a unified action space; reviewers validate schemas and inspect a sample of trajectories to confirm executable fidelity. Models trained on FineVision consistently outperform those trained on existing open mixtures across a broad evaluation suite, underscoring the benefits of scale, data hygiene, and balanced automation with human oversight. We release the corpus and curation tools to accelerate data-centric VLM research.

Post-training pruning can substantially reduce LLM inference costs, but it often degrades quality unless the remaining weights are adapted. Since global retraining is expensive at LLM scale, recent work has largely focused on increasingly sophisticated pruning criteria that aim to select better sparsity patterns without adaptation. We revisit this trade-off through local reconstruction: after pruning, we adapt one subset of the model parameters at a time on a calibration set, training it to match the corresponding intermediate activations of the dense model. We evaluate local reconstruction across model families and scales, up to 72B parameters, and establish three main findings. First, local reconstruction is an effective adaptation mechanism for LLMs: it matches post-pruning retraining while using over an order of magnitude less data and compute, even when using PEFT techniques. Second, reconstruction exhibits a broad "free-lunch" regime in granularity, i.e., the reconstruction parameter window: as long as the reconstructed region contains at least a nonlinear submodule, final quality is largely insensitive to the window size, allowing granularity to be chosen primarily based on memory constraints. In contrast, reconstructing individual matrices, despite being the natural approach often proposed in the literature, consistently underperforms, as small matrix-level errors accumulate into larger activation drift. Lastly, reconstruction reduces the relative importance of the pruning criterion: performance gaps between sophisticated criteria and simple baselines shrink with model scale, making simple methods competitive again. Overall, our results challenge the prevailing view that post-pruning adaptation is impractical for LLMs.

The output of image the segmentation process is usually not very clear due to low quality features of Satellite images. The purpose of this study is to find a suitable Conditional Random Field (CRF) to achieve better clarity in a segmented image. We started with different types of CRFs and studied them as to why they are or are not suitable for our purpose. We evaluated our approach on two different datasets - Satellite imagery having low quality features and high quality Aerial photographs. During the study we experimented with various CRFs to find which CRF gives the best results on images and compared our results on these datasets to show the pitfalls and potentials of different approaches.

Iconicity, the resemblance between linguistic form and meaning, is pervasive in signed languages, offering a natural testbed for visual grounding. For vision-language models (VLMs), the challenge is to recover such essential mappings from dynamic human motion rather than static context. We introduce the Visual Iconicity Challenge, a novel video-based benchmark that adapts psycholinguistic measures to evaluate VLMs on three tasks: (i) phonological sign-form prediction (e.g., handshape, location), (ii) transparency (inferring meaning from visual form), and (iii) graded iconicity ratings. We assess 13 state-of-the-art VLMs in zero- and few-shot settings on Sign Language of the Netherlands and compare them to human baselines. On phonological form prediction, VLMs recover some handshape and location detail but remain below human performance; on transparency, they are far from human baselines; and only top models correlate moderately with human iconicity ratings. Interestingly, models with stronger phonological form prediction correlate better with human iconicity judgment, indicating shared sensitivity to visually grounded structure. Our findings validate these diagnostic tasks and motivate human-centric signals and embodied learning methods for modelling iconicity and improving visual grounding in multimodal models.

To drive progress in science and engineering, large language models (LLMs) must be able to process large amounts of numerical data and solve long calculations efficiently. This is currently only possible through the use of external tools or extensive reasoning chains, either weakening the numerical representations of LLMs or limiting the length of problems they can solve. We show that frontier LLMs require excessive amounts of reasoning tokens to solve even basic calculations, which is exacerbated by their tokenization strategies that split single numbers into multiple tokens. This motivates the need for efficient and effective single-token number encodings. We introduce a set of desiderata for such encodings and show that existing approaches fail to fulfill them. To address these shortcomings, we propose BitTokens, a novel encoding strategy that represents any number as a single token using its IEEE 754 binary floating-point representation. Through extensive experiments we show that our BitTokens allow even small language models to learn algorithms that solve basic arithmetic operations nearly perfectly. This newly gained efficiency could expand the length and complexity of problems language models can solve.

Foundation models are reshaping medical imaging, yet their application in echocardiography remains limited, hindered by a heavy reliance on private datasets that prevent reproducible comparison. Echocardiography poses unique challenges, including noisy acquisitions, high frame redundancy, and limited diverse public datasets. To address this, we introduce CardioBench, a comprehensive benchmark for echocardiography foundation models. Specifically, CardioBench unifies eight publicly available datasets into a standardized suite spanning four regression and five classification tasks, covering functional, structural, diagnostic, and view recognition endpoints. Leveraging this framework, we evaluate several leading foundation models, including cardiac-specific, biomedical, and general-purpose encoders, under consistent zero-shot, probing, and alignment protocols. Our analysis reveals that while general-purpose encoders transfer well and often close the gap with probing, they struggle significantly with fine-grained distinctions like view classification and subtle pathology recognition. Results indicate that models capturing temporal cardiac dynamics perform best on functional tasks, while retrieval-based approaches generalize more consistently across datasets. By releasing preprocessing, splits, and public evaluation pipelines, CardioBench establishes a reproducible reference point to guide the architectural design of future echocardiography and possibly other medical imaging foundation models.

Designing dense rewards is crucial for reinforcement learning (RL), yet in robotics it often demands extensive manual effort and lacks scalability. One promising solution is to view task progress as a dense reward signal, as it quantifies the degree to which actions advance the system toward task completion over time. We present TimeRewarder, a simple yet effective reward learning method that derives progress estimation signals from passive videos, including robot demonstrations and human videos, by modeling temporal distances between frame pairs. We then demonstrate how TimeRewarder can supply step-wise proxy rewards to guide reinforcement learning. In our comprehensive experiments on ten challenging Meta-World tasks, we show that TimeRewarder dramatically improves RL for sparse-reward tasks, achieving nearly perfect success in 9/10 tasks with only 200,000 environment interactions per task. This approach outperformed previous methods and even the manually designed environment dense reward on both the final success rate and sample efficiency. Moreover, we show that TimeRewarder pretraining can exploit real-world human videos, highlighting its potential as a scalable approach to rich reward signals from diverse video sources.

This article initiates the study of a basic question about model pruning. Given a vector $s$ of importance scores assigned to model components, how many of the scored components could be discarded without sacrificing performance? We propose Effective Model Pruning (EMP), which derives the desired sparsity directly from the score distribution using the notion of effective sample size from particle filtering, also known as the inverse Simpson index. Rather than prescribe a pruning criterion, EMP supplies a universal adaptive threshold derived from the distribution of the score $s$ over the model components: EMP maps $s$ to a number $N_{eff}=N_{eff}(s)$, called the effective sample size. The $N-N_{eff}$ lowest scoring components are discarded. A tight lower bound on the effective mass $s_{eff}$ (the sum of retained normalized scores) in terms of $N_{eff}$ is derived. This process yields models with a provable upper bound on the loss change relative to the original dense model. Numerical experiments are performed demonstrating this phenomenon across a variety of network architectures including MLPs, CNNs, Transformers, LLMs, and KAN. It is also shown that EMP addresses a rich set of pruning criteria such as weight magnitude, attention score, KAN importance score, and even feature-level signals such as image pixels.

Reinforcement learning (RL) struggles to scale to large, combinatorial action spaces common in many real-world problems. This paper introduces a novel framework for training discrete diffusion models as highly effective policies in these complex settings. Our key innovation is an efficient online training process that ensures stable and effective policy improvement. By leveraging policy mirror descent (PMD) to define an ideal, regularized target policy distribution, we frame the policy update as a distributional matching problem, training the expressive diffusion model to replicate this stable target. This decoupled approach stabilizes learning and significantly enhances training performance. Our method achieves state-of-the-art results and superior sample efficiency across a diverse set of challenging combinatorial benchmarks, including DNA sequence generation, RL with macro-actions, and multi-agent systems. Experiments demonstrate that our diffusion policies attain superior performance compared to other baselines.

Accurate multi-needle localization in intraoperative CT images is crucial for optimizing seed placement in pelvic seed implant brachytherapy. However, this task is challenging due to poor image contrast and needle adhesion. This paper presents a novel approach that reframes needle localization as a tip-handle detection and matching problem to overcome these difficulties. An anchor-free network, based on HRNet, is proposed to extract multi-scale features and accurately detect needle tips and handles by predicting their centers and orientations using decoupled branches for heatmap regression and polar angle prediction. To associate detected tips and handles into individual needles, a greedy matching and merging (GMM) method designed to solve the unbalanced assignment problem with constraints (UAP-C) is presented. The GMM method iteratively selects the most probable tip-handle pairs and merges them based on a distance metric to reconstruct 3D needle paths. Evaluated on a dataset of 100 patients, the proposed method demonstrates superior performance, achieving higher precision and F1 score compared to a segmentation-based method utilizing the nnUNet model,thereby offering a more robust and accurate solution for needle localization in complex clinical scenarios.

Vision Transformers (ViTs) achieve state-of-the-art performance in semantic segmentation but are hindered by high computational and memory costs. To address this, we propose STEP (SuperToken and Early-Pruning), a hybrid token-reduction framework that combines dynamic patch merging and token pruning to enhance efficiency without significantly compromising accuracy. At the core of STEP is dCTS, a lightweight CNN-based policy network that enables flexible merging into superpatches. Encoder blocks integrate also early-exits to remove high-confident supertokens, lowering computational load. We evaluate our method on high-resolution semantic segmentation benchmarks, including images up to 1024 x 1024, and show that when dCTS is applied alone, the token count can be reduced by a factor of 2.5 compared to the standard 16 x 16 pixel patching scheme. This yields a 2.6x reduction in computational cost and a 3.4x increase in throughput when using ViT-Large as the backbone. Applying the full STEP framework further improves efficiency, reaching up to a 4x reduction in computational complexity and a 1.7x gain in inference speed, with a maximum accuracy drop of no more than 2.0%. With the proposed STEP configurations, up to 40% of tokens can be confidently predicted and halted before reaching the final encoder layer.

Generative recommendation plays a crucial role in personalized systems, predicting users' future interactions from their historical behavior sequences. A critical yet underexplored factor in training these models is data augmentation, the process of constructing training data from user interaction histories. By shaping the training distribution, data augmentation directly and often substantially affects model generalization and performance. Nevertheless, in much of the existing work, this process is simplified, applied inconsistently, or treated as a minor design choice, without a systematic and principled understanding of its effects. Motivated by our empirical finding that different augmentation strategies can yield large performance disparities, we conduct an in-depth analysis of how they reshape training distributions and influence alignment with future targets and generalization to unseen inputs. To systematize this design space, we propose GenPAS, a generalized and principled framework that models augmentation as a stochastic sampling process over input-target pairs with three bias-controlled steps: sequence sampling, target sampling, and input sampling. This formulation unifies widely used strategies as special cases and enables flexible control of the resulting training distribution. Our extensive experiments on benchmark and industrial datasets demonstrate that GenPAS yields superior accuracy, data efficiency, and parameter efficiency compared to existing strategies, providing practical guidance for principled training data construction in generative recommendation. Our code is available at https://github.com/snap-research/GenPAS.

3D Gaussian Splatting (3DGS) is rapidly gaining popularity for its photorealistic rendering quality and real-time performance, but it generates massive amounts of data. Hence compressing 3DGS data is necessary for the cost effectiveness of 3DGS models. Recently, several anchor-based neural compression methods have been proposed, achieving good 3DGS compression performance. However, they all rely on uniform scalar quantization (USQ) due to its simplicity. A tantalizing question is whether more sophisticated quantizers can improve the current 3DGS compression methods with very little extra overhead and minimal change to the system. The answer is yes by replacing USQ with lattice vector quantization (LVQ). To better capture scene-specific characteristics, we optimize the lattice basis for each scene, improving LVQ's adaptability and R-D efficiency. This scene-adaptive LVQ (SALVQ) strikes a balance between the R-D efficiency of vector quantization and the low complexity of USQ. SALVQ can be seamlessly integrated into existing 3DGS compression architectures, enhancing their R-D performance with minimal modifications and computational overhead. Moreover, by scaling the lattice basis vectors, SALVQ can dynamically adjust lattice density, enabling a single model to accommodate multiple bit rate targets. This flexibility eliminates the need to train separate models for different compression levels, significantly reducing training time and memory consumption.

Multi-Target Multi-Camera Tracking (MTMC) is an essential computer vision task for automating large-scale surveillance. With camera calibration and depth information, the targets in the scene can be projected into 3D space, offering unparalleled levels of automatic perception of a 3D environment. However, tracking in the 3D space requires replacing all 2D tracking components from the ground up, which may be infeasible for existing MTMC systems. In this paper, we present an approach for extending any online 2D multi-camera tracking system into 3D space by utilizing depth information to reconstruct a target in point-cloud space, and recovering its 3D box through clustering and yaw refinement following tracking. We also introduced an enhanced online data association mechanism that leverages the target's local ID consistency to assign global IDs across frames. The proposed framework is evaluated on the 2025 AI City Challenge's 3D MTMC dataset, achieving 3rd place on the leaderboard.

Large language models (LLMs)-empowered autonomous agents are transforming both digital and physical environments by enabling adaptive, multi-agent collaboration. While these agents offer significant opportunities across domains such as finance, healthcare, and smart manufacturing, their unpredictable behaviors and heterogeneous capabilities pose substantial governance and accountability challenges. In this paper, we propose a blockchain-enabled layered architecture for regulatory agent collaboration, comprising an agent layer, a blockchain data layer, and a regulatory application layer. Within this framework, we design three key modules: (i) an agent behavior tracing and arbitration module for automated accountability, (ii) a dynamic reputation evaluation module for trust assessment in collaborative scenarios, and (iii) a malicious behavior forecasting module for early detection of adversarial activities. Our approach establishes a systematic foundation for trustworthy, resilient, and scalable regulatory mechanisms in large-scale agent ecosystems. Finally, we discuss the future research directions for blockchain-enabled regulatory frameworks in multi-agent systems.

Explainability, in particular, the ability for robots to explain why they have made a decision or behaved in a certain way, is a critical tool in helping users understand the robots they interact and coexist with. Behaviour trees are a popular framework for controlling the decision-making of robots, and thus a natural question to ask is whether or not a system driven by a behaviour tree is capable of answering "why" questions. While explainability for behaviour tree-driven robots has seen some prior attention, no existing methods are capable of generating causal, counterfactual explanations which detail the reasons for robot decisions and behaviour. Therefore, in this work, we introduce a novel approach which automatically generates counterfactual explanations in response to contrastive "why" questions. Our method achieves this by first automatically building a causal model from the structure of the behaviour tree as well as domain knowledge about the state and individual behaviour tree nodes. The resultant causal model is then queried and searched to find a set of diverse counterfactual explanations. We demonstrate that our approach is able to correctly explain the behaviour of a wide range of behaviour tree structures and states in real time, unlike previous methods which are either unable to answer contrastive questions with causal explanations, or are not guaranteed to provide consistent and accurate explanations. By being able to answer a wide range of causal queries, our approach represents a step towards more transparent, understandable, and ultimately safe and trustworthy robotic systems.

The recent advancements of Large Language Models (LLMs) have spurred considerable research interest in extending their linguistic capabilities beyond text to other modalities, which leads to emergence of speech-based LLMs (SpeechLMs) with capability of processing user request in either speech or textual formats. However, owing to inter-modal discrepancies, these SpeechLMs still exhibit a significant performance gap compared to their text-based LLM counterparts in instruction-following, particularly when confronted with the dynamic and variable nature of user speech. To address this challenge, this paper introduces a framework termed Reinforced Behavior Alignment (RBA), designed to bolster the language generation proficiency of SpeechLMs. Instead of relying on supervised fine-tuning from human annotations, RBA employs a self-synthesis methodology to generate extensive, high-fidelity alignment data by a powerful teacher LLM. Then SpeechLMs is aligned its behavior with that of a teacher using a reinforcement learning-based approach. Experimental results demonstrate that this method effectively enhances the instruction-following capabilities of SpeechLMs that outperform conventional distillation baselines. Crucially, we demonstrate that RBA can be seamlessly extended to tasks such including spoken question answering and speech-to-text translation, attaining state-of-the-art performance on open benchmarks with only self-generated data.

Purpose: This study introduces the first adaptation of RETFound for joint optic disc (OD) and optic cup (OC) segmentation. RETFound is a well-known foundation model developed for fundus camera and optical coherence tomography images, which has shown promising performance in disease diagnosis. Methods: We propose FunduSegmenter, a model integrating a series of novel modules with RETFound, including a Pre-adapter, a Decoder, a Post-adapter, skip connections with Convolutional Block Attention Module and a Vision Transformer block adapter. The model is evaluated on a proprietary dataset, GoDARTS, and four public datasets, IDRiD, Drishti-GS, RIM-ONE-r3, and REFUGE, through internal verification, external verification and domain generalization experiments. Results: An average Dice similarity coefficient of 90.51% was achieved in internal verification, which outperformed all baselines, some substantially (nnU-Net: 82.91%; DUNet: 89.17%; TransUNet: 87.91%). In all external verification experiments, the average results were about 3% higher than those of the best baseline, and our model was also competitive in domain generalization. Conclusions: This study explored the potential of the latent general representations learned by RETFound for OD and OC segmentation in fundus camera images. Our FunduSegmenter generally outperformed state-of-the-art baseline methods. The proposed modules are general and can be extended to fine-tuning other foundation models. Translational Relevance: The model shows strong stability and generalization on both in-distribution and out-of-distribution data, providing stable OD and OC segmentation. This is an essential step for many automated tasks, from setting the accurate retinal coordinate to biomarker discovery. The code and trained weights are available at: https://github.com/JusticeZzy/FunduSegmenter.

Large Language Models (LLMs) are increasingly deployed in long-context tasks such as reasoning, code generation, and multi-turn dialogue. However, inference over extended contexts is bottlenecked by the Key-Value (KV) cache, whose memory footprint grows linearly with sequence length and dominates latency at each decoding step. While recent KV cache compression methods identify and load important few tokens, they focus predominantly on input contexts and fail to address the cumulative attention errors that arise during long decoding. In this paper, we introduce RetroAttention, a novel KV cache update technique that retrospectively revises past attention outputs using newly arrived KV entries from subsequent decoding steps. By maintaining a lightweight output cache, RetroAttention enables past queries to be efficiently supplemented with more contexts, while incurring minimal latency overhead. This breaks the fixed-attention-output paradigm and allows continual correction of prior approximations. Extensive experiments on long-generation benchmarks show that RetroAttention consistently outperforms state-of-the-art (SOTA) KV compression methods, increasing effective KV exposure by up to 1.6$\times$ and accuracy by up to 21.9\%.

Large language models (LLMs) have revolutionized artificial intelligence by enabling complex reasoning capabilities. While recent advancements in reinforcement learning (RL) have primarily focused on domain-specific reasoning tasks (e.g., mathematics or code generation), real-world reasoning scenarios often require models to handle diverse and complex environments that narrow-domain benchmarks cannot fully capture. To address this gap, we present InternBootcamp, an open-source framework comprising 1000+ domain-diverse task environments specifically designed for LLM reasoning research. Our codebase offers two key functionalities: (1) automated generation of unlimited training/testing cases with configurable difficulty levels, and (2) integrated verification modules for objective response evaluation. These features make InternBootcamp fundamental infrastructure for RL-based model optimization, synthetic data generation, and model evaluation. Although manually developing such a framework with enormous task coverage is extremely cumbersome, we accelerate the development procedure through an automated agent workflow supplemented by manual validation protocols, which enables the task scope to expand rapidly. % With these bootcamps, we further establish Bootcamp-EVAL, an automatically generated benchmark for comprehensive performance assessment. Evaluation reveals that frontier models still underperform in many reasoning tasks, while training with InternBootcamp provides an effective way to significantly improve performance, leading to our 32B model that achieves state-of-the-art results on Bootcamp-EVAL and excels on other established benchmarks. In particular, we validate that consistent performance gains come from including more training tasks, namely \textbf{task scaling}, over two orders of magnitude, offering a promising route towards capable reasoning generalist.

Affordance grounding focuses on predicting the specific regions of objects that are associated with the actions to be performed by robots. It plays a vital role in the fields of human-robot interaction, human-object interaction, embodied manipulation, and embodied perception. Existing models often neglect the affordance shared among different objects because they lack the Chain-of-Thought(CoT) reasoning abilities, limiting their out-of-domain (OOD) generalization and explicit reasoning capabilities. To address these challenges, we propose Affordance-R1, the first unified affordance grounding framework that integrates cognitive CoT guided Group Relative Policy Optimization (GRPO) within a reinforcement learning paradigm. Specifically, we designed a sophisticated affordance function, which contains format, perception, and cognition rewards to effectively guide optimization directions. Furthermore, we constructed a high-quality affordance-centric reasoning dataset, ReasonAff, to support training. Trained exclusively via reinforcement learning with GRPO and without explicit reasoning data, Affordance-R1 achieves robust zero-shot generalization and exhibits emergent test-time reasoning capabilities. Comprehensive experiments demonstrate that our model outperforms well-established methods and exhibits open-world generalization. To the best of our knowledge, Affordance-R1 is the first to integrate GRPO-based RL with reasoning into affordance reasoning. The code of our method and our dataset is released on https://github.com/hq-King/Affordance-R1.

Multimodal sentiment analysis (MSA) aims to understand human emotions by integrating information from multiple modalities, such as text, audio, and visual data. However, existing methods often suffer from spurious correlations both within and across modalities, leading models to rely on statistical shortcuts rather than true causal relationships, thereby undermining generalization. To mitigate this issue, we propose a Multi-relational Multimodal Causal Intervention (MMCI) framework, which leverages the backdoor adjustment from causal theory to address the confounding effects of such shortcuts. Specifically, we first model the multimodal inputs as a multi-relational graph to explicitly capture intra- and inter-modal dependencies. Then, we apply an attention mechanism to separately estimate and disentangle the causal features and shortcut features corresponding to these intra- and inter-modal relations. Finally, by applying the backdoor adjustment, we stratify the shortcut features and dynamically combine them with the causal features to encourage MMCI to produce stable predictions under distribution shifts. Extensive experiments on several standard MSA datasets and out-of-distribution (OOD) test sets demonstrate that our method effectively suppresses biases and improves performance.

Radar-based human pose estimation enables privacy-preserving motion tracking for ambient intelligence, yet the noisy nature of radar sensing makes uncertainty quantification essential. We present RadProPoser, an end-to-end probabilistic framework that predicts three-dimensional body joints with per-joint uncertainties from raw radar tensor data. Using a variational encoder-decoder with spectral attention that fuses real and imaginary radar components across temporal frames, we model aleatoric uncertainty through learnable Gaussian and Laplace distributions. Trained on a new benchmark dataset with optical motion-capture ground truth, our method achieves 6.425 cm mean per-joint position error. The model outputs per-joint aleatoric uncertainties, and isotonic recalibration yields calibrated total uncertainty with expected calibration error of 0.027. Since spectral attention operates on individual radar tensor components, extending to multi-radar configurations requires only concatenating additional input streams. On the HuPR benchmark with dual orthogonal radars, this achieves 5.042 cm MPJPE. The framework runs at 89 frames per second (FPS) on an NVIDIA RTX 3090, exceeding the 15 Hz radar frame rate.

Structured pruning is a standard tool for compressing deep neural networks, but its practical performance depends on how sparsity is allocated across layers. We propose FAIR-Pruner, a search-free framework for adaptive layer-wise structured pruning. FAIR-Pruner uses two within-layer rankings: a removal-oriented signal that proposes candidate units and a protection-oriented signal that identifies task-sensitive units. Its core component, Tolerance of Difference (ToD), measures the overlap between the removal prefix and the protected tail, and uses a shared tolerance level to induce non-uniform pruning depths across layers. As a default vision instantiation, FAIR-Pruner combines a Wasserstein-based U-Score for class-conditional unit separability with a Taylor-based R-Score for task-level sensitivity; the same ToD allocation rule can also be paired with alternative removal signals. Theoretically, we analyze ToD through the population R-Score, derive rank-based control of the high-R-Score mass entering the pruning set, and identify an additive exchange condition for same-budget comparison with uniform pruning. Experiments on CIFAR-10, CIFAR-100, SVHN, and ImageNet across VGG, ResNet, DenseNet, ConvNeXt, and DeiT show strong accuracy--compression trade-offs. Prune-only experiments on routed-expert Qwen1.5-MoE-A2.7B-Chat further examine architectural extensibility under matched expert budgets. FAIR-Pruner is released as a pip-installable open-source package.

Text-to-image diffusion models have demonstrated remarkable capabilities in generating artistic content by learning from billions of images, including popular artworks. However, the fundamental question of how these models internally represent concepts, such as content and style in paintings, remains unexplored. Traditional computer vision assumes content and style are orthogonal, but diffusion models receive no explicit guidance about this distinction during training. In this work, we investigate how transformer-based text-to-image diffusion models encode content and style concepts when generating artworks. We leverage cross-attention heatmaps to attribute pixels in generated images to specific prompt tokens, enabling us to isolate image regions influenced by content-describing versus style-describing tokens. Our findings reveal that diffusion models demonstrate varying degrees of content-style separation depending on the specific artistic prompt and style requested. In many cases, content tokens primarily influence object-related regions while style tokens affect background and texture areas, suggesting an emergent understanding of the content-style distinction. These insights contribute to our understanding of how large-scale generative models internally represent complex artistic concepts without explicit supervision. We share the code and dataset, together with an exploratory tool for visualizing attention maps at https://github.com/umilISLab/artistic-prompt-interpretation.

Depth information is robust to scene appearance variations and inherently carries 3D spatial details. Thus, a visual backbone based on the vision transformer is proposed to fuse RGB and depth modalities for enhancing generalization in this paper. Different modalities are first processed by separate CNN stems, and the combined convolutional features are delivered to the scalable vision transformer to obtain visual representations. Moreover, a contrastive learning scheme is designed with masked and unmasked tokens to enhance the sample efficiency and generalization performance. A curriculum-based domain randomization scheme is used to flexibly stabilize the training process. Finally, simulation results demonstrate that our fusion scheme outperforms the other baselines. The feasibility of our model is validated to perform real-world manipulation tasks via zero-shot transfer.

Long-horizon goal-conditioned tasks pose fundamental challenges for reinforcement learning (RL), particularly when goals are distant and rewards are sparse. While hierarchical and graph-based methods offer partial solutions, their reliance on conventional hindsight relabeling often fails to correct subgoal infeasibility, leading to inefficient high-level planning. To address this, we propose Strict Subgoal Execution (SSE), a graph-based hierarchical RL framework that integrates Frontier Experience Replay (FER) to separate unreachable from admissible subgoals and streamline high-level decision making. FER delineates the reachability frontier using failure and partial-success transitions, which identifies unreliable subgoals, increases subgoal reliability, and reduces unnecessary high-level decisions. Additionally, SSE employs a decoupled exploration policy to cover underexplored regions of the goal space and a path refinement that adjusts edge costs using observed low-level failures. Experimental results across diverse long-horizon benchmarks show that SSE consistently outperforms existing goal-conditioned and hierarchical RL methods in both efficiency and success rate. Our code is available at https://jaebak1996.github.io/SSE/

Time series forecasting aims to model temporal dependencies among variables for future state inference, holding significant importance and widespread applications in real-world scenarios. Although deep learning-based methods have achieved remarkable progress, they still exhibit suboptimal performance in long-term forecasting. Recent research demonstrates that large language models (LLMs) achieve promising performance in time series forecasting, but this progress is still met with skepticism about whether LLMs are truly useful for this task. To address this, we propose Time-Prompt, a framework for activating LLMs for time series forecasting. Specifically, we first construct a unified prompt paradigm with learnable soft prompts to guide the LLM's behavior and textualized hard prompts to enhance the time series representations. Second, to enhance LLM' comprehensive understanding of the forecasting task, we design a semantic space embedding and cross-modal alignment module to achieve fusion of temporal and textual data. Finally, we efficiently fine-tune the LLM's parameters using time series data. Furthermore, we focus on carbon emissions, aiming to provide a modest contribution to global carbon neutrality. Comprehensive evaluations on 6 public datasets and 3 carbon emission datasets demonstrate that Time-Prompt is a powerful framework for time series forecasting.

Large Language Models (LLMs) have demonstrated remarkable general capabilities, but enhancing skills such as reasoning often demands substantial computational resources and may compromise generalization. While Parameter-Efficient Fine-Tuning (PEFT) methods offer a more resource-conscious alternative, they typically require retraining for each LLM backbone due to architectural dependencies. To address these challenges, we propose Universal Reasoner (UniR)-a modular, composable, and plug-and-play reasoning module that can be used with larger frozen LLMs to provide specialized reasoning capabilities with a shared or aligned token space. Specifically, UniR decomposes the reward into a standalone reasoning module trained in a decoupled manner using verifiable rewards, effectively translating trajectory-level signals into token-level guidance. Once trained, UniR is combined with frozen LLMs at inference time by simply adding its output logits to those of the backbone. This additive structure enables modular composition: multiple UniR modules trained for different tasks can be jointly applied by summing their logits, enabling complex reasoning via composition. Furthermore, UniR demonstrates weak-to-strong generalization, where reasoning modules trained on smaller models effectively guide much larger LLMs in the same model family, and generalize across domains such as in vision language models and medical reasoning. Experiments on mathematical reasoning and machine translation show that UniR surpasses existing fine-tuning methods. Code is open-sourced at https://github.com/hangeol/UniR.

Chatbots via large language models (LLMs) generate fluent responses but often struggle with when to speak, especially for brief, timely listener reactions during ongoing dialogue. We present a multimodal strategy for LLMs, which leverages synchronized video, audio, and text cues to improve conversational timing awareness. The strategy reformulates response timing as a dense response-type prediction task, enabling an agent to decide whether to remain silent, produce a short reaction, or start a full response under streaming constraints. Therefore, we introduce a curated multimodal dataset from real-world dyadic conversational videos with temporally aligned modalities and fine-grained reaction type annotations. Moreover, we design a multimodal strategy, MM-When2Speak, with a multimodal integration module on top of an LLM backbone. Experiments across various modality settings and strong LLM baselines show that MM-When2Speak achieves up to a 3x improvement in response type prediction performance, highlighting the importance of multimodal perception for natural and engaging conversational interaction.