Jingtao Xu

Zizhuo Lin, Quanling Liu, Jinsheng Quan et al.

Large language models (LLMs) often solve a task when all instructions are given in a single prompt, but fail when the same information is revealed gradually across turns. When a clean FULL prompt and a RAW-SHARDED conversation contain the same complete user evidence, the model should still arrive at the same answer. We argue that a key reason for this gap is self-anchored drift: responses produced under partial information introduce unsupported assumptions, and those assumptions later distort the final answer. To reduce this effect, we propose Canonical-Context On-Policy Distillation (CCOPD). During training, the same base model is used in two roles: a frozen teacher conditioned on the clean FULL prompt and a trainable student that receives the same evidence incrementally through a multi-turn conversation; CCOPD aligns the student's behavior on its own trajectories with the teacher's canonical full-context behavior. Trained only on math problem conversations, CCOPD yields a 32\% average relative improvement in RAW-SHARDED performance over the original base model across math and five zero-shot out-of-domain task families, while largely preserving full-context performance. Further analyses suggest that CCOPD strengthens grounding in user evidence and reduces sensitivity to contamination from earlier assistant turns.

Guibin Chen, Dixuan Lin, Jiangping Yang et al.

SkyReels V4 is a unified multi modal video foundation model for joint video audio generation, inpainting, and editing. The model adopts a dual stream Multimodal Diffusion Transformer (MMDiT) architecture, where one branch synthesizes video and the other generates temporally aligned audio, while sharing a powerful text encoder based on the Multimodal Large Language Models (MMLM). SkyReels V4 accepts rich multi modal instructions, including text, images, video clips, masks, and audio references. By combining the MMLMs multi modal instruction following capability with in context learning in the video branch MMDiT, the model can inject fine grained visual guidance under complex conditioning, while the audio branch MMDiT simultaneously leverages audio references to guide sound generation. On the video side, we adopt a channel concatenation formulation that unifies a wide range of inpainting style tasks, such as image to video, video extension, and video editing under a single interface, and naturally extends to vision referenced inpainting and editing via multi modal prompts. SkyReels V4 supports up to 1080p resolution, 32 FPS, and 15 second duration, enabling high fidelity, multi shot, cinema level video generation with synchronized audio. To make such high resolution, long-duration generation computationally feasible, we introduce an efficiency strategy: Joint generation of low resolution full sequences and high-resolution keyframes, followed by dedicated super-resolution and frame interpolation models. To our knowledge, SkyReels V4 is the first video foundation model that simultaneously supports multi-modal input, joint video audio generation, and a unified treatment of generation, inpainting, and editing, while maintaining strong efficiency and quality at cinematic resolutions and durations.

Zhengcong Fei, Debang Li, Di Qiu et al.

This paper presents SkyReels-A2, a controllable video generation framework capable of assembling arbitrary visual elements (e.g., characters, objects, backgrounds) into synthesized videos based on textual prompts while maintaining strict consistency with reference images for each element. We term this task elements-to-video (E2V), whose primary challenges lie in preserving the fidelity of each reference element, ensuring coherent composition of the scene, and achieving natural outputs. To address these, we first design a comprehensive data pipeline to construct prompt-reference-video triplets for model training. Next, we propose a novel image-text joint embedding model to inject multi-element representations into the generative process, balancing element-specific consistency with global coherence and text alignment. We also optimize the inference pipeline for both speed and output stability. Moreover, we introduce a carefully curated benchmark for systematic evaluation, i.e, A2 Bench. Experiments demonstrate that our framework can generate diverse, high-quality videos with precise element control. SkyReels-A2 is the first open-source commercial grade model for the generation of E2V, performing favorably against advanced closed-source commercial models. We anticipate SkyReels-A2 will advance creative applications such as drama and virtual e-commerce, pushing the boundaries of controllable video generation.

Hanqi Feng, Peng Qiu, Mengchun Zhang et al.

Recent advances in diffusion models have shown remarkable potential for antibody design, yet existing approaches apply uniform generation strategies that cannot adapt to each antigen's unique requirements. Inspired by B cell affinity maturation, where antibodies evolve through multi-objective optimization balancing affinity, stability, and self-avoidance, we propose the first biologically-motivated framework that leverages physics-based domain knowledge within an online meta-learning system. Our method employs multiple specialized experts (van der Waals, molecular recognition, energy balance, and interface geometry) whose parameters evolve during generation based on iterative feedback, mimicking natural antibody refinement cycles. Instead of fixed protocols, this adaptive guidance discovers personalized optimization strategies for each target. Our experiments demonstrate that this approach: (1) discovers optimal SE(3)-equivariant guidance strategies for different antigen classes without pre-training, preserving molecular symmetries throughout optimization; (2) significantly enhances hotspot coverage and interface quality through target-specific adaptation, achieving balanced multi-objective optimization characteristic of therapeutic antibodies; (3) establishes a paradigm for iterative refinement where each antibody-antigen system learns its unique optimization profile through online evaluation; (4) generalizes effectively across diverse design challenges, from small epitopes to large protein interfaces, enabling precision-focused campaigns for individual targets.

Xueqiao Zhang, Chao Zhang, Jingtao Xu et al.

Role-playing agents (RPAs) have attracted growing interest for their ability to simulate immersive and interactive characters. However, existing approaches primarily focus on static role profiles, overlooking the dynamic perceptual abilities inherent to humans. To bridge this gap, we introduce the concept of dynamic role profiles by incorporating video modality into RPAs. To support this, we construct Role-playing-Video60k, a large-scale, high-quality dataset comprising 60k videos and 700k corresponding dialogues. Based on this dataset, we develop a comprehensive RPA framework that combines adaptive temporal sampling with both dynamic and static role profile representations. Specifically, the dynamic profile is created by adaptively sampling video frames and feeding them to the LLM in temporal order, while the static profile consists of (1) character dialogues from training videos during fine-tuning, and (2) a summary context from the input video during inference. This joint integration enables RPAs to generate greater responses. Furthermore, we propose a robust evaluation method covering eight metrics. Experimental results demonstrate the effectiveness of our framework, highlighting the importance of dynamic role profiles in developing RPAs.

Jingtao Xu, Yali Li, Shengjin Wang

Detection in large-scale scenes is a challenging problem due to small objects and extreme scale variation. It is essential to focus on the image regions of small objects. In this paper, we propose a novel Adaptive Zoom (AdaZoom) network as a selective magnifier with flexible shape and focal length to adaptively zoom the focus regions for object detection. Based on policy gradient, we construct a reinforcement learning framework for focus region generation, with the reward formulated by object distributions. The scales and aspect ratios of the generated regions are adaptive to the scales and distribution of objects inside. We apply variable magnification according to the scale of the region for adaptive multi-scale detection. We further propose collaborative training to complementarily promote the performance of AdaZoom and the detection network. To validate the effectiveness, we conduct extensive experiments on VisDrone2019, UAVDT, and DOTA datasets. The experiments show AdaZoom brings a consistent and significant improvement over different detection networks, achieving state-of-the-art performance on these datasets, especially outperforming the existing methods by AP of 4.64% on Vis-Drone2019.