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.

StereoWorld: Geometry-Aware Monocular-to-Stereo Video Generation

The growing adoption of XR devices has fueled strong demand for high-quality stereo video, yet its production remains costly and artifact-prone. To address this challenge, we present StereoWorld, an end-to-end framework that repurposes a pretrained video generator for high-fidelity monocular-to-stereo video generation. Our framework jointly conditions the model on the monocular video input while explicitly supervising the generation with a geometry-aware regularization to ensure 3D structural fidelity. A spatio-temporal tiling scheme is further integrated to enable efficient, high-resolution synthesis. To enable large-scale training and evaluation, we curate a high-definition stereo video dataset containing over 11M frames aligned to natural human interpupillary distance (IPD). Extensive experiments demonstrate that StereoWorld substantially outperforms prior methods, generating stereo videos with superior visual fidelity and geometric consistency. The project webpage is available at https://ke-xing.github.io/StereoWorld/.

Beat the long tail: Distribution-Aware Speculative Decoding for RL Training

Reinforcement learning(RL) post-training has become essential for aligning large language models (LLMs), yet its efficiency is increasingly constrained by the rollout phase, where long trajectories are generated token by token. We identify a major bottleneck:the long-tail distribution of rollout lengths, where a small fraction of long generations dominates wall clock time and a complementary opportunity; the availability of historical rollouts that reveal stable prompt level patterns across training epochs. Motivated by these observations, we propose DAS, a Distribution Aware Speculative decoding framework that accelerates RL rollouts without altering model outputs. DAS integrates two key ideas: an adaptive, nonparametric drafter built from recent rollouts using an incrementally maintained suffix tree, and a length aware speculation policy that allocates more aggressive draft budgets to long trajectories that dominate makespan. This design exploits rollout history to sustain acceptance while balancing base and token level costs during decoding. Experiments on math and code reasoning tasks show that DAS reduces rollout time up to 50% while preserving identical training curves, demonstrating that distribution-aware speculative decoding can significantly accelerate RL post training without compromising learning quality.

Maritime Communication in Evaporation Duct Environment with Ship Trajectory Optimization

In maritime wireless networks, the evaporation duct effect has been known as a preferable condition for long-range transmissions. However, how to effectively utilize the duct effect for efficient communication design is still open for investigation. In this paper, we consider a typical scenario of ship-to-shore data transmission, where a ship collects data from multiple oceanographic buoys, sails from one to another, and transmits the collected data back to a terrestrial base station during its voyage. A novel framework, which exploits priori information of the channel gain map in the presence of evaporation duct, is proposed to minimize the data transmission time and the sailing time by optimizing the ship's trajectory. To this end, a multi-objective optimization problem is formulated and is further solved by a dynamic population PSO-integrated NSGA-II algorithm. Through simulations, it is demonstrated that, compared to the benchmark scheme which ignores useful information of the evaporation duct, the proposed scheme can effectively reduce both the data transmission time and the sailing time.

Staircase Streaming for Low-Latency Multi-Agent Inference

Recent advances in large language models (LLMs) opened up new directions for leveraging the collective expertise of multiple LLMs. These methods, such as Mixture-of-Agents, typically employ additional inference steps to generate intermediate outputs, which are then used to produce the final response. While multi-agent inference can enhance response quality, it can significantly increase the time to first token (TTFT), posing a challenge for latency-sensitive applications and hurting user experience. To address this issue, we propose staircase streaming for low-latency multi-agent inference. Instead of waiting for the complete intermediate outputs from previous steps, we begin generating the final response as soon as we receive partial outputs from these steps. Experimental results demonstrate that staircase streaming reduces TTFT by up to 93% while maintaining response quality.

MOOSE-Chem3: Toward Experiment-Guided Hypothesis Ranking via Simulated Experimental Feedback

Hypothesis ranking is a crucial component of automated scientific discovery, particularly in natural sciences where wet-lab experiments are costly and throughput-limited. Existing approaches focus on pre-experiment ranking, relying solely on large language model's internal reasoning without incorporating empirical outcomes from experiments. We introduce the task of experiment-guided ranking, which aims to prioritize candidate hypotheses based on the results of previously tested ones. However, developing such strategies is challenging due to the impracticality of repeatedly conducting real experiments in natural science domains. To address this, we propose a simulator grounded in three domain-informed assumptions, modeling hypothesis performance as a function of similarity to a known ground truth hypothesis, perturbed by noise. We curate a dataset of 124 chemistry hypotheses with experimentally reported outcomes to validate the simulator. Building on this simulator, we develop a pseudo experiment-guided ranking method that clusters hypotheses by shared functional characteristics and prioritizes candidates based on insights derived from simulated experimental feedback. Experiments show that our method outperforms pre-experiment baselines and strong ablations.

FloE: On-the-Fly MoE Inference on Memory-constrained GPU

With the widespread adoption of Mixture-of-Experts (MoE) models, there is a growing demand for efficient inference on memory-constrained devices. While offloading expert parameters to CPU memory and loading activated experts on demand has emerged as a potential solution, the large size of activated experts overburdens the limited PCIe bandwidth, hindering the effectiveness in latency-sensitive scenarios. To mitigate this, we propose FloE, an on-the-fly MoE inference system on memory-constrained GPUs. FloE is built on the insight that there exists substantial untapped redundancy within sparsely activated experts. It employs various compression techniques on the expert's internal parameter matrices to reduce the data movement load, combined with low-cost sparse prediction, achieving perceptible inference acceleration in wall-clock time on resource-constrained devices. Empirically, FloE achieves a 9.3x compression of parameters per expert in Mixtral-8x7B; enables deployment on a GPU with only 11GB VRAM, reducing the memory footprint by up to 8.5x; and delivers a 48.7x inference speedup compared to DeepSpeed-MII on a single GeForce RTX 3090 - all with only a 4.4$\%$ - 7.6$\%$ average performance degradation.

FloE: On-the-Fly MoE Inference

With the widespread adoption of Mixture-of-Experts (MoE) models, there is a growing demand for efficient inference on memory-constrained devices. While offloading expert parameters to CPU memory and loading activated experts on demand has emerged as a potential solution, the large size of activated experts overburdens the limited PCIe bandwidth, hindering the effectiveness in latency-sensitive scenarios. To mitigate this, we propose FloE, an on-the-fly MoE inference system on memory-constrained GPUs. FloE is built on the insight that there exists substantial untapped redundancy within sparsely activated experts. It employs various compression techniques on the expert's internal parameter matrices to reduce the data movement load, combined with low-cost sparse prediction, achieving perceptible inference acceleration in wall-clock time on resource-constrained devices. Empirically, FloE achieves a 9.3x compression of parameters per expert in Mixtral-8x7B; enables deployment on a GPU with only 11GB VRAM, reducing the memory footprint by up to 8.5x; and delivers a 48.7x inference speedup compared to DeepSpeed-MII on a single GeForce RTX 3090.