EXaMCaP: Subset Selection with Entropy Gain Maximization for Probing Capability Gains of Large Chart Understanding Training Sets

Recent works focus on synthesizing Chart Understanding (ChartU) training sets to inject advanced chart knowledge into Multimodal Large Language Models (MLLMs), where the sufficiency of the knowledge is typically verified by quantifying capability gains via the fine-tune-then-evaluate paradigm. However, full-set fine-tuning MLLMs to assess such gains incurs significant time costs, hindering the iterative refinement cycles of the ChartU dataset. Reviewing the ChartU dataset synthesis and data selection domains, we find that subsets can potentially probe the MLLMs' capability gains from full-set fine-tuning. Given that data diversity is vital for boosting MLLMs' performance and entropy reflects this feature, we propose EXaMCaP, which uses entropy gain maximization to select a subset. To obtain a high-diversity subset, EXaMCaP chooses the maximum-entropy subset from the large ChartU dataset. As enumerating all possible subsets is impractical, EXaMCaP iteratively selects samples to maximize the gain in set entropy relative to the current set, approximating the maximum-entropy subset of the full dataset. Experiments show that EXaMCaP outperforms baselines in probing the capability gains of the ChartU training set, along with its strong effectiveness across diverse subset sizes and compatibility with various MLLM architectures.

LongVPO: From Anchored Cues to Self-Reasoning for Long-Form Video Preference Optimization

We present LongVPO, a novel two-stage Direct Preference Optimization framework that enables short-context vision-language models to robustly understand ultra-long videos without any long-video annotations. In Stage 1, we synthesize preference triples by anchoring questions to individual short clips, interleaving them with distractors, and applying visual-similarity and question-specificity filtering to mitigate positional bias and ensure unambiguous supervision. We also approximate the reference model's scoring over long contexts by evaluating only the anchor clip, reducing computational overhead. In Stage 2, we employ a recursive captioning pipeline on long videos to generate scene-level metadata, then use a large language model to craft multi-segment reasoning queries and dispreferred responses, aligning the model's preferences through multi-segment reasoning tasks. With only 16K synthetic examples and no costly human labels, LongVPO outperforms the state-of-the-art open-source models on multiple long-video benchmarks, while maintaining strong short-video performance (e.g., on MVBench), offering a scalable paradigm for efficient long-form video understanding.

DC-VLAQ: Query-Residual Aggregation for Robust Visual Place Recognition

One of the central challenges in visual place recognition (VPR) is learning a robust global representation that remains discriminative under large viewpoint changes, illumination variations, and severe domain shifts. While visual foundation models (VFMs) provide strong local features, most existing methods rely on a single model, overlooking the complementary cues offered by different VFMs. However, exploiting such complementary information inevitably alters token distributions, which challenges the stability of existing query-based global aggregation schemes. To address these challenges, we propose DC-VLAQ, a representation-centric framework that integrates the fusion of complementary VFMs and robust global aggregation. Specifically, we first introduce a lightweight residual-guided complementary fusion that anchors representations in the DINOv2 feature space while injecting complementary semantics from CLIP through a learned residual correction. In addition, we propose the Vector of Local Aggregated Queries (VLAQ), a query--residual global aggregation scheme that encodes local tokens by their residual responses to learnable queries, resulting in improved stability and the preservation of fine-grained discriminative cues. Extensive experiments on standard VPR benchmarks, including Pitts30k, Tokyo24/7, MSLS, Nordland, SPED, and AmsterTime, demonstrate that DC-VLAQ consistently outperforms strong baselines and achieves state-of-the-art performance, particularly under challenging domain shifts and long-term appearance changes.

STaR: Sensitive Trajectory Regulation for Unlearning in Large Reasoning Models

Large Reasoning Models (LRMs) have advanced automated multi-step reasoning, but their ability to generate complex Chain-of-Thought (CoT) trajectories introduces severe privacy risks, as sensitive information may be deeply embedded throughout the reasoning process. Existing Large Language Models (LLMs) unlearning approaches that typically focus on modifying only final answers are insufficient for LRMs, as they fail to remove sensitive content from intermediate steps, leading to persistent privacy leakage and degraded security. To address these challenges, we propose Sensitive Trajectory Regulation (STaR), a parameter-free, inference-time unlearning framework that achieves robust privacy protection throughout the reasoning process. Specifically, we first identify sensitive content via semantic-aware detection. Then, we inject global safety constraints through secure prompt prefix. Next, we perform trajectory-aware suppression to dynamically block sensitive content across the entire reasoning chain. Finally, we apply token-level adaptive filtering to prevent both exact and paraphrased sensitive tokens during generation. Furthermore, to overcome the inadequacies of existing evaluation protocols, we introduce two metrics: Multi-Decoding Consistency Assessment (MCS), which measures the consistency of unlearning across diverse decoding strategies, and Multi-Granularity Membership Inference Attack (MIA) Evaluation, which quantifies privacy protection at both answer and reasoning-chain levels. Experiments on the R-TOFU benchmark demonstrate that STaR achieves comprehensive and stable unlearning with minimal utility loss, setting a new standard for privacy-preserving reasoning in LRMs.

Seedance 1.5 pro: A Native Audio-Visual Joint Generation Foundation Model

Recent strides in video generation have paved the way for unified audio-visual generation. In this work, we present Seedance 1.5 pro, a foundational model engineered specifically for native, joint audio-video generation. Leveraging a dual-branch Diffusion Transformer architecture, the model integrates a cross-modal joint module with a specialized multi-stage data pipeline, achieving exceptional audio-visual synchronization and superior generation quality. To ensure practical utility, we implement meticulous post-training optimizations, including Supervised Fine-Tuning (SFT) on high-quality datasets and Reinforcement Learning from Human Feedback (RLHF) with multi-dimensional reward models. Furthermore, we introduce an acceleration framework that boosts inference speed by over 10X. Seedance 1.5 pro distinguishes itself through precise multilingual and dialect lip-syncing, dynamic cinematic camera control, and enhanced narrative coherence, positioning it as a robust engine for professional-grade content creation. Seedance 1.5 pro is now accessible on Volcano Engine at https://console.volcengine.com/ark/region:ark+cn-beijing/experience/vision?type=GenVideo.

Adaptive Frequency Domain Alignment Network for Medical image segmentation

High-quality annotated data plays a crucial role in achieving accurate segmentation. However, such data for medical image segmentation are often scarce due to the time-consuming and labor-intensive nature of manual annotation. To address this challenge, we propose the Adaptive Frequency Domain Alignment Network (AFDAN)--a novel domain adaptation framework designed to align features in the frequency domain and alleviate data scarcity. AFDAN integrates three core components to enable robust cross-domain knowledge transfer: an Adversarial Domain Learning Module that transfers features from the source to the target domain; a Source-Target Frequency Fusion Module that blends frequency representations across domains; and a Spatial-Frequency Integration Module that combines both frequency and spatial features to further enhance segmentation accuracy across domains. Extensive experiments demonstrate the effectiveness of AFDAN: it achieves an Intersection over Union (IoU) of 90.9% for vitiligo segmentation in the newly constructed VITILIGO2025 dataset and a competitive IoU of 82.6% on the retinal vessel segmentation benchmark DRIVE, surpassing existing state-of-the-art approaches.

RewardDance: Reward Scaling in Visual Generation

Reward Models (RMs) are critical for improving generation models via Reinforcement Learning (RL), yet the RM scaling paradigm in visual generation remains largely unexplored. It primarily due to fundamental limitations in existing approaches: CLIP-based RMs suffer from architectural and input modality constraints, while prevalent Bradley-Terry losses are fundamentally misaligned with the next-token prediction mechanism of Vision-Language Models (VLMs), hindering effective scaling. More critically, the RLHF optimization process is plagued by Reward Hacking issue, where models exploit flaws in the reward signal without improving true quality. To address these challenges, we introduce RewardDance, a scalable reward modeling framework that overcomes these barriers through a novel generative reward paradigm. By reformulating the reward score as the model's probability of predicting a "yes" token, indicating that the generated image outperforms a reference image according to specific criteria, RewardDance intrinsically aligns reward objectives with VLM architectures. This alignment unlocks scaling across two dimensions: (1) Model Scaling: Systematic scaling of RMs up to 26 billion parameters; (2) Context Scaling: Integration of task-specific instructions, reference examples, and chain-of-thought (CoT) reasoning. Extensive experiments demonstrate that RewardDance significantly surpasses state-of-the-art methods in text-to-image, text-to-video, and image-to-video generation. Crucially, we resolve the persistent challenge of "reward hacking": Our large-scale RMs exhibit and maintain high reward variance during RL fine-tuning, proving their resistance to hacking and ability to produce diverse, high-quality outputs. It greatly relieves the mode collapse problem that plagues smaller models.

PAE MobiLLM: Privacy-Aware and Efficient LLM Fine-Tuning on the Mobile Device via Additive Side-Tuning

There is a huge gap between numerous intriguing applications fostered by on-device large language model (LLM) fine-tuning (FT) from fresh mobile data and the limited resources of a mobile device. While existing server-assisted methods (e.g., split learning or side-tuning) may enable LLM FT on the local mobile device, they suffer from heavy communication burdens of activation transmissions, and may disclose data, labels or fine-tuned models to the server. To address those issues, we develop PAE MobiLLM, a privacy-aware and efficient LLM FT method which can be deployed on the mobile device via server-assisted additive side-tuning. To further accelerate FT convergence and improve computing efficiency, PAE MobiLLM integrates activation caching on the server side, which allows the server to reuse historical activations and saves the mobile device from repeatedly computing forward passes for the recurring data samples. Besides, to reduce communication cost, PAE MobiLLM develops a one-token (i.e., ``pivot'' token) activation shortcut that transmits only a single activation dimension instead of full activation matrices to guide the side network tuning. Last but not least, PAE MobiLLM introduces the additive adapter side-network design which makes the server train the adapter modules based on device-defined prediction differences rather than raw ground-truth labels. In this way, the server can only assist device-defined side-network computing, and learn nothing about data, labels or fine-tuned models.

Seedance 1.0: Exploring the Boundaries of Video Generation Models

Notable breakthroughs in diffusion modeling have propelled rapid improvements in video generation, yet current foundational model still face critical challenges in simultaneously balancing prompt following, motion plausibility, and visual quality. In this report, we introduce Seedance 1.0, a high-performance and inference-efficient video foundation generation model that integrates several core technical improvements: (i) multi-source data curation augmented with precision and meaningful video captioning, enabling comprehensive learning across diverse scenarios; (ii) an efficient architecture design with proposed training paradigm, which allows for natively supporting multi-shot generation and jointly learning of both text-to-video and image-to-video tasks. (iii) carefully-optimized post-training approaches leveraging fine-grained supervised fine-tuning, and video-specific RLHF with multi-dimensional reward mechanisms for comprehensive performance improvements; (iv) excellent model acceleration achieving ~10x inference speedup through multi-stage distillation strategies and system-level optimizations. Seedance 1.0 can generate a 5-second video at 1080p resolution only with 41.4 seconds (NVIDIA-L20). Compared to state-of-the-art video generation models, Seedance 1.0 stands out with high-quality and fast video generation having superior spatiotemporal fluidity with structural stability, precise instruction adherence in complex multi-subject contexts, native multi-shot narrative coherence with consistent subject representation.

Budget-Adaptive Adapter Tuning in Orthogonal Subspaces for Continual Learning in LLMs

Large language models (LLMs) often suffer from catastrophic forgetting in continual learning (CL) scenarios, where performance on previously learned tasks degrades severely while training on sequentially arriving tasks. Although pioneering CL approaches using orthogonal subspaces can mitigate task interference, they typically employ fixed budget allocation, neglecting the varying complexity across tasks and layers. Besides, recent budget-adaptive tuning methods for LLMs often adopt multi-stage paradigms that decouple optimization and budget allocation. Such decoupling results in potential misalignment, which hinders those approaches' practical application in CL scenarios. To address these limitations, we propose OA-Adapter, a novel parameter-efficient approach for continual learning in LLMs that unifies dynamic budget adaptation with orthogonal subspace learning in a single end-to-end training stage. Specifically, OA-Adapter introduces a dynamic bottleneck dimension adaptation mechanism that simultaneously allocates an efficient parameter budget and optimizes task objectives without misalignment. To effectively preserve previously acquired knowledge while coordinating with the dynamic budget allocation, orthogonal constraints are applied specifically between the parameter subspace of the current task and the dynamically allocated parameter subspaces of historical tasks. Experimental results on continual learning benchmarks demonstrate that OA-Adapter outperforms state-of-the-art methods in both accuracy and parameter efficiency, achieving higher average accuracy while using 58.5% fewer parameters on the standard CL benchmark.