SAW-INT4: System-Aware 4-Bit KV-Cache Quantization for Real-World LLM Serving

KV-cache memory is a major bottleneck in real-world LLM serving, where systems must simultaneously support latency-sensitive small-batch requests and high-throughput concurrent workloads. Although many KV-cache compression methods improve offline accuracy or compression ratio, they often violate practical serving constraints such as paged memory layouts, regular memory access, and fused attention execution, limiting their effectiveness in deployment. In this work, we identify the minimal set of 4-bit KV-cache quantization methods that remain viable under these constraints. Our central finding is that a simple design--token-wise INT4 quantization with block-diagonal Hadamard rotation--consistently achieves the best accuracy-efficiency trade-off. Across multiple models and benchmarks, this approach recovers nearly all of the accuracy lost by naive INT4, while more complex methods such as vector quantization and Hessian-aware quantization provide only marginal additional gains once serving compatibility is taken into account. To make this practical, we implement a fused rotation-quantization kernel that integrates directly into paged KV-cache layouts and introduces zero measurable end-to-end overhead, matching plain INT4 throughput across concurrency levels. Our results show that effective KV-cache compression is fundamentally a systems co-design problem: under real serving constraints, lightweight block-diagonal Hadamard rotation is a viable method that delivers near-lossless accuracy without sacrificing serving efficiency.

Invertible Diffusion for Low-Memory Channel Gain Map Construction in Wireless Communication Networks

Channel gain maps (CGMs) enable propagation-aware services in edge-intelligent wireless communication networks, while diffusion-based CGM construction is memory intensive for on-device training or adaptation. This letter proposes InvDiff-CGM, an invertible diffusion framework that constructs CGMs from sparse measurements and environmental priors. By adopting invertible architectures in both the diffusion process and the U-Net noise estimator, InvDiff-CGM achieves near-constant training memory consumption. A prior-informed multi-scale injector further integrates environmental priors with sparse measurements to improve physical consistency and detail preservation. Experiments on RadioMap3DSeer show about an 85\% reduction in peak training memory and a PSNR of 38.02~dB, outperforming representative recent baselines. This validates the practicality of InvDiff-CGM for high-fidelity CGM construction under edge resource constraints.

STORM: End-to-End Referring Multi-Object Tracking in Videos

Referring multi-object tracking (RMOT) is a task of associating all the objects in a video that semantically match with given textual queries or referring expressions. Existing RMOT approaches decompose object grounding and tracking into separated modules and exhibit limited performance due to the scarcity of training videos, ambiguous annotations, and restricted domains. In this work, we introduce STORM, an end-to-end MLLM that jointly performs grounding and tracking within a unified framework, eliminating external detectors and enabling coherent reasoning over appearance, motion, and language. To improve data efficiency, we propose a task-composition learning (TCL) strategy that decomposes RMOT into image grounding and object tracking, allowing STORM to leverage data-rich sub-tasks and learn structured spatial--temporal reasoning. We further construct STORM-Bench, a new RMOT dataset with accurate trajectories and diverse, unambiguous referring expressions generated through a bottom-up annotation pipeline. Extensive experiments show that STORM achieves state-of-the-art performance on image grounding, single-object tracking, and RMOT benchmarks, demonstrating strong generalization and robust spatial--temporal grounding in complex real-world scenarios. STORM-Bench is released at https://github.com/amazon-science/storm-referring-multi-object-grounding.

Squeeze Evolve: Unified Multi-Model Orchestration for Verifier-Free Evolution

We show that verifier-free evolution is bottlenecked by both diversity and efficiency: without external correction, repeated evolution accelerates collapse toward narrow modes, while the uniform use of a high-cost model wastes compute and quickly becomes economically impractical. We introduce Squeeze Evolve, a unified multi-model orchestration framework for verifier-free evolutionary inference. Our approach is guided by a simple principle: allocate model capability where it has the highest marginal utility. Stronger models are reserved for high-impact stages, while cheaper models handle the other stages at much lower costs. This principle addresses diversity and cost-efficiency jointly while remaining lightweight. Squeeze Evolve naturally supports open-source, closed-source, and mixed-model deployments. Across AIME 2025, HMMT 2025, LiveCodeBench V6, GPQA-Diamond, ARC-AGI-V2, and multimodal vision benchmarks, such as MMMU-Pro and BabyVision, Squeeze Evolve consistently improves the cost-capability frontier over single-model evolution and achieves new state-of-the-art results on several tasks. Empirically, Squeeze Evolve reduces API cost by up to $\sim$3$\times$ and increases fixed-budget serving throughput by up to $\sim$10$\times$. Moreover, on discovery tasks, Squeeze Evolve is the first verifier-free evolutionary method to match, and in some cases exceed, the performance of verifier-based evolutionary methods.

PolyReal: A Benchmark for Real-World Polymer Science Workflows

Multimodal Large Language Models (MLLMs) excel in general domains but struggle with complex, real-world science. We posit that polymer science, an interdisciplinary field spanning chemistry, physics, biology, and engineering, is an ideal high-stakes testbed due to its diverse multimodal data. Yet, existing benchmarks related to polymer science largely overlook real-world workflows, limiting their practical utility and failing to systematically evaluate MLLMs across the full, practice-grounded lifecycle of experimentation. We introduce PolyReal, a novel multimodal benchmark grounded in real-world scientific practices to evaluate MLLMs on the full lifecycle of polymer experimentation. It covers five critical capabilities: (1) foundational knowledge application; (2) lab safety analysis; (3) experiment mechanism reasoning; (4) raw data extraction; and (5) performance & application exploration. Our evaluation of leading MLLMs on PolyReal reveals a capability imbalance. While models perform well on knowledge-intensive reasoning (e.g., Experiment Mechanism Reasoning), they drop sharply on practice-based tasks (e.g., Lab Safety Analysis and Raw Data Extraction). This exposes a severe gap between abstract scientific knowledge and its practical, context-dependent application, showing that these real-world tasks remain challenging for MLLMs. Thus, PolyReal helps address this evaluation gap and provides a practical benchmark for assessing AI systems in real-world scientific workflows.

OpenDPR: Open-Vocabulary Change Detection via Vision-Centric Diffusion-Guided Prototype Retrieval for Remote Sensing Imagery

Open-vocabulary change detection (OVCD) seeks to recognize arbitrary changes of interest by enabling generalization beyond a fixed set of predefined classes. We reformulate OVCD as a two-stage pipeline: first generate class-agnostic change proposals using visual foundation models (VFMs) such as SAM and DINOv2, and then perform category identification with vision-language models (VLMs) such as CLIP. We reveal that category identification errors are the primary bottleneck of OVCD, mainly due to the limited ability of VLMs based on image-text matching to represent fine-grained land-cover categories. To address this, we propose OpenDPR, a training-free vision-centric diffusion-guided prototype retrieval framework. OpenDPR leverages diffusion models to construct diverse prototypes for target categories offline, and to perform similarity retrieval with change proposals in the visual space during inference. The secondary bottleneck lies in change localization, due to the inherent lack of change priors in VFMs. To bridge this gap, we design a spatial-to-change weakly supervised change detection module named S2C to adapt their strong spatial modeling capabilities for change localization. Integrating the pretrained S2C into OpenDPR leads to an optional weakly supervised variant named OpenDPR-W, which further improves OVCD with minimal supervision. Experimental results on four benchmark datasets demonstrate that the proposed methods achieve state-of-the-art performance under both supervision modes. Code is available at https://github.com/guoqi2002/OpenDPR.

A Unified Codebook Design for Curvature-Reconfigurable Apertures: Seamless Near to Far Field Coverage

Beam training for extremely large-scale arrays with curvature-reconfigurable apertures (CuRAs) faces the critical challenge of severe, geometry-dependent angle-range coupling. While most existing designs compartmentalize near field and far field scenarios, we propose a unified, distance-adaptive hierarchical codebook framework for 1-D and 2-D CuRAs that seamlessly bridges both propagation regimes. Under a spherical-wave model, we first characterize the beamforming-gain correlation in a polar angular domain, deriving an angle-dependent angular sampling rule to capture the varying curvature. To achieve full-range coverage, we introduce a direction-dependent effective Rayleigh distance (ERD) as a soft boundary to gate the range sampling. Crucially, by sampling uniformly in the reciprocal-range domain, the proposed codebook provides precise, dense focusing within the ERD and automatically degenerates into sparse, angle-only steering beyond it. This mechanism eliminates the need for hard mode-switching between near- and far-field operations. Simulation results demonstrate that our unified design consistently outperforms representative baselines in spectral efficiency and alignment accuracy, offering a comprehensive solution for full-range CuRA communications.

Magic-MM-Embedding: Towards Visual-Token-Efficient Universal Multimodal Embedding with MLLMs

Multimodal Large Language Models (MLLMs) have shown immense promise in universal multimodal retrieval, which aims to find relevant items of various modalities for a given query. But their practical application is often hindered by the substantial computational cost incurred from processing a large number of tokens from visual inputs. In this paper, we propose Magic-MM-Embedding, a series of novel models that achieve both high efficiency and state-of-the-art performance in universal multimodal embedding. Our approach is built on two synergistic pillars: (1) a highly efficient MLLM architecture incorporating visual token compression to drastically reduce inference latency and memory footprint, and (2) a multi-stage progressive training strategy designed to not only recover but significantly boost performance. This coarse-to-fine training paradigm begins with extensive continue pretraining to restore multimodal understanding and generation capabilities, progresses to large-scale contrastive pretraining and hard negative mining to enhance discriminative power, and culminates in a task-aware fine-tuning stage guided by an MLLM-as-a-Judge for precise data curation. Comprehensive experiments show that our model outperforms existing methods by a large margin while being more inference-efficient.

Understanding and Steering the Cognitive Behaviors of Reasoning Models at Test-Time

Large Language Models (LLMs) often rely on long chain-of-thought (CoT) reasoning to solve complex tasks. While effective, these trajectories are frequently inefficient, leading to high latency from excessive token generation, or unstable reasoning that alternates between underthinking (shallow, inconsistent steps) and overthinking (repetitive, verbose reasoning). In this work, we study the structure of reasoning trajectories and uncover specialized attention heads that correlate with distinct cognitive behaviors such as verification and backtracking. By lightly intervening on these heads at inference time, we can steer the model away from inefficient modes. Building on this insight, we propose CREST, a training-free method for Cognitive REasoning Steering at Test-time. CREST has two components: (1) an offline calibration step that identifies cognitive heads and derives head-specific steering vectors, and (2) an inference-time procedure that rotates hidden representations to suppress components along those vectors. CREST adaptively suppresses unproductive reasoning behaviors, yielding both higher accuracy and lower computational cost. Across diverse reasoning benchmarks and models, CREST improves accuracy by up to 17.5% while reducing token usage by 37.6%, offering a simple and effective pathway to faster, more reliable LLM reasoning.

PersonaLive! Expressive Portrait Image Animation for Live Streaming

Current diffusion-based portrait animation models predominantly focus on enhancing visual quality and expression realism, while overlooking generation latency and real-time performance, which restricts their application range in the live streaming scenario. We propose PersonaLive, a novel diffusion-based framework towards streaming real-time portrait animation with multi-stage training recipes. Specifically, we first adopt hybrid implicit signals, namely implicit facial representations and 3D implicit keypoints, to achieve expressive image-level motion control. Then, a fewer-step appearance distillation strategy is proposed to eliminate appearance redundancy in the denoising process, greatly improving inference efficiency. Finally, we introduce an autoregressive micro-chunk streaming generation paradigm equipped with a sliding training strategy and a historical keyframe mechanism to enable low-latency and stable long-term video generation. Extensive experiments demonstrate that PersonaLive achieves state-of-the-art performance with up to 7-22x speedup over prior diffusion-based portrait animation models.