Trending Papers

SAM 3D is a generative model that reconstructs 3D objects from single images using a multi-stage training framework that includes synthetic pretraining and real-world alignment, achieving high performance in human preference tests.

Today's denoising diffusion models do not "denoise" in the classical sense, i.e., they do not directly predict clean images. Rather, the neural networks predict noise or a noised quantity. In this paper, we suggest that predicting clean data and predicting noised quantities are fundamentally different. According to the manifold assumption, natural data should lie on a low-dimensional manifold, whereas noised quantities do not. With this assumption, we advocate for models that directly predict clean data, which allows apparently under-capacity networks to operate effectively in very high-dimensional spaces. We show that simple, large-patch Transformers on pixels can be strong generative models: using no tokenizer, no pre-training, and no extra loss. Our approach is conceptually nothing more than "Just image Transformers", or JiT, as we call it. We report competitive results using JiT with large patch sizes of 16 and 32 on ImageNet at resolutions of 256 and 512, where predicting high-dimensional noised quantities can fail catastrophically. With our networks mapping back to the basics of the manifold, our research goes back to basics and pursues a self-contained paradigm for Transformer-based diffusion on raw natural data.

Depth Anything 3 (DA3) uses a plain transformer for geometry prediction from visual inputs, achieving state-of-the-art results in camera pose estimation, any-view geometry, visual rendering, and monocular depth estimation.

MHR combines ATLAS's skeleton/shape paradigm with a modern rig to provide expressive, anatomically plausible human animation for AR/VR and graphics.

LightRAG improves Retrieval-Augmented Generation by integrating graph structures for enhanced contextual awareness and efficient information retrieval, achieving better accuracy and response times.

3D modeling is shifting from static visual representations toward physical, articulated assets that can be directly used in simulation and interaction. However, most existing 3D generation methods overlook key physical and articulation properties, thereby limiting their utility in embodied AI. To bridge this gap, we introduce PhysX-Anything, the first simulation-ready physical 3D generative framework that, given a single in-the-wild image, produces high-quality sim-ready 3D assets with explicit geometry, articulation, and physical attributes. Specifically, we propose the first VLM-based physical 3D generative model, along with a new 3D representation that efficiently tokenizes geometry. It reduces the number of tokens by 193x, enabling explicit geometry learning within standard VLM token budgets without introducing any special tokens during fine-tuning and significantly improving generative quality. In addition, to overcome the limited diversity of existing physical 3D datasets, we construct a new dataset, PhysX-Mobility, which expands the object categories in prior physical 3D datasets by over 2x and includes more than 2K common real-world objects with rich physical annotations. Extensive experiments on PhysX-Mobility and in-the-wild images demonstrate that PhysX-Anything delivers strong generative performance and robust generalization. Furthermore, simulation-based experiments in a MuJoCo-style environment validate that our sim-ready assets can be directly used for contact-rich robotic policy learning. We believe PhysX-Anything can substantially empower a broad range of downstream applications, especially in embodied AI and physics-based simulation.

PaddleOCR-VL, a vision-language model combining NaViT-style visual encoder and ERNIE-4.5 language model, achieves state-of-the-art performance in document parsing with minimal resource consumption.

Flawed-Aware Policy Optimization (FAPO) enhances reinforcement learning with verifiable rewards by penalizing flawed-positive rollouts, improving reasoning capability and training stability in large language models.

MiMo-Embodied, a cross-embodied foundation model, achieves state-of-the-art performance in both autonomous driving and embodied AI through multi-stage learning, curated data, and CoT/RL fine-tuning.

NaTex generates 3D textures directly using latent color diffusion and geometry-aware models, outperforming previous methods in coherence and alignment.

We present MiroThinker v1.0, an open-source research agent designed to advance tool-augmented reasoning and information-seeking capabilities. Unlike previous agents that only scale up model size or context length, MiroThinker explores interaction scaling at the model level, systematically training the model to handle deeper and more frequent agent-environment interactions as a third dimension of performance improvement. Unlike LLM test-time scaling, which operates in isolation and risks degradation with longer reasoning chains, interactive scaling leverages environment feedback and external information acquisition to correct errors and refine trajectories. Through reinforcement learning, the model achieves efficient interaction scaling: with a 256K context window, it can perform up to 600 tool calls per task, enabling sustained multi-turn reasoning and complex real-world research workflows. Across four representative benchmarks-GAIA, HLE, BrowseComp, and BrowseComp-ZH-the 72B variant achieves up to 81.9%, 37.7%, 47.1%, and 55.6% accuracy respectively, surpassing previous open-source agents and approaching commercial counterparts such as GPT-5-high. Our analysis reveals that MiroThinker benefits from interactive scaling consistently: research performance improves predictably as the model engages in deeper and more frequent agent-environment interactions, demonstrating that interaction depth exhibits scaling behaviors analogous to model size and context length. These findings establish interaction scaling as a third critical dimension for building next-generation open research agents, complementing model capacity and context windows.

Step-Audio-R1, using the Modality-Grounded Reasoning Distillation framework, achieves strong reasoning capabilities in audio, outperforming previous models and demonstrating the transferability of reasoning across modalities.

Kandinsky 5.0 is a family of state-of-the-art generative models for high-resolution images and short videos, featuring model lineups with varying parameters and enhanced training techniques to achieve superior quality and performance.

LlamaFactory is a unified framework enabling efficient fine-tuning of large language models across various tasks using a web-based user interface.

Agent Lightning is a flexible RL framework for training LLMs in various agents, using a hierarchical RL algorithm and decoupling execution from training to handle complex interactions.

A multi-agent framework using large language models for stock trading simulates real-world trading firms, improving performance metrics like cumulative returns and Sharpe ratio.

IndexTTS, an enhanced text-to-speech system combining XTTS and Tortoise models, offers improved naturalness, enhanced voice cloning, and controllable usage through hybrid character-pinyin modeling and optimized vector quantization.

A novel GPT-based model, OmniFlatten, enables real-time natural full-duplex spoken dialogue through a multi-stage post-training technique that integrates speech and text without altering the original model's architecture.

DeepAnalyze-8B, an agentic LLM, autonomously completes the data science pipeline from raw data to research reports using curriculum-based training and data-grounded trajectory synthesis.

MinerU2.5, a 1.2B-parameter document parsing vision-language model, achieves state-of-the-art recognition accuracy with computational efficiency through a coarse-to-fine parsing strategy.

MinerU is an open-source tool that enhances document content extraction using fine-tuned models and pre/postprocessing rules across diverse document types.

VANS, a model combining reinforcement learning, a Vision-Language Model, and a Video Diffusion Model, achieves state-of-the-art performance in Video-Next-Event Prediction by generating visually consistent and semantically accurate videos.

A parallel multimodal diffusion framework, MMaDA-Parallel, enhances cross-modal alignment and semantic consistency in thinking-aware image synthesis by addressing error propagation issues in sequential approaches.

Mem0, a memory-centric architecture with graph-based memory, enhances long-term conversational coherence in LLMs by efficiently extracting, consolidating, and retrieving information, outperforming existing memory systems in terms of accuracy and computational efficiency.

Agent0, a self-evolving framework utilizing multi-step co-evolution and tool integration, enhances LLM reasoning capabilities without human-curated data.

MIRIX, a modular multi-agent memory system, enhances language models' memory capabilities by integrating diverse memory types and a dynamic framework, achieving superior performance in multimodal and long-form conversation benchmarks.

We present Uni-MoE 2.0 from the Lychee family. As a fully open-source omnimodal large model (OLM), it substantially advances Lychee's Uni-MoE series in language-centric multimodal understanding, reasoning, and generating. Based on the Qwen2.5-7B dense architecture, we build Uni-MoE-2.0-Omni from scratch through three core contributions: dynamic-capacity Mixture-of-Experts (MoE) design, a progressive training strategy enhanced with an iterative reinforcement strategy, and a carefully curated multimodal data matching technique. It is capable of omnimodal understanding, as well as generating images, text, and speech. Architecturally, our new MoE framework balances computational efficiency and capability for 10 cross-modal inputs using shared, routed, and null experts, while our Omni-Modality 3D RoPE ensures spatio-temporal cross-modality alignment in the self-attention layer. For training, following cross-modal pretraining, we use a progressive supervised fine-tuning strategy that activates modality-specific experts and is enhanced by balanced data composition and an iterative GSPO-DPO method to stabilise RL training and improve reasoning. Data-wise, the base model, trained on approximately 75B tokens of open-source multimodal data, is equipped with special speech and image generation tokens, allowing it to learn these generative tasks by conditioning its outputs on linguistic cues. Extensive evaluation across 85 benchmarks demonstrates that our model achieves SOTA or highly competitive performance against leading OLMs, surpassing Qwen2.5-Omni (trained with 1.2T tokens) on over 50 of 76 benchmarks. Key strengths include video understanding (+7% avg. of 8), omnimodallity understanding (+7% avg. of 4), and audiovisual reasoning (+4%). It also advances long-form speech processing (reducing WER by 4.2%) and leads in low-level image processing and controllable generation across 5 metrics.

InternVL3 is a multimodal pre-trained language model that jointly learns from both multimodal data and text, improving performance and scalability through advanced techniques and setting a new state-of-the-art in multimodal tasks.

A novel method, CoTyle, generates images in consistent visual styles using unique numerical style codes, filling an academic gap in code-to-style image generation.

VibeThinker-1.5B, a 1.5B-parameter model using the Spectrum-to-Signal Principle, achieves superior reasoning capabilities compared to larger models at a significantly lower cost.

PyTorch Fully Sharded Data Parallel (FSDP) enables efficient and scalable training of large models across hardware configurations.

The PyTorch distributed data parallel module optimizes large-scale model training using techniques like gradient bucketing, computation-communication overlap, and selective synchronization to achieve near-linear scalability.

Zep, a memory layer service, outperforms MemGPT in the DMR benchmark and LongMemEval by excelling in dynamic knowledge integration and temporal reasoning, critical for enterprise use cases.

UniMoE-Audio, a unified speech and music generation model using a Dynamic-Capacity Mixture-of-Experts framework, addresses data imbalance and task conflicts, achieving state-of-the-art performance and enhanced cross-domain synergy.

A sparse Multimodal Mixture of Experts (MoE) model, Uni-MoE, effectively handles multiple data types with efficient training and improved performance through modality-specific encoders, cross-modality alignment, and Low-Rank Adaptation.

DINOv3, a self-supervised learning model, achieves superior performance across various vision tasks by scaling datasets and models, addressing dense feature degradation, and enhancing flexibility with post-hoc strategies.

RAG-Anything is a unified framework that enhances multimodal knowledge retrieval by integrating cross-modal relationships and semantic matching, outperforming existing methods on complex benchmarks.

V-ReasonBench evaluates generative video models across structured problem-solving, spatial cognition, pattern-based inference, and physical dynamics using a diverse set of tasks.

MediaPipe framework facilitates the development of perception applications by providing tools for combining components, prototyping, and measuring performance across platforms.

AgentScope enhances agentic applications by providing flexible tool-based interactions, unified interfaces, and advanced infrastructure based on the ReAct paradigm, supporting efficient and safe development and deployment.

A framework called Easy Dataset synthesizes fine-tuning data from unstructured documents using a GUI and LLMs, improving domain-specific performance of LLMs while maintaining general knowledge.

MemOS, a memory operating system for Large Language Models, addresses memory management challenges by unifying plaintext, activation-based, and parameter-level memories, enabling efficient storage, retrieval, and continual learning.

A method using 3D Gaussians for scene representation and optimized rendering allows high-quality, real-time novel-view synthesis at 1080p resolution.

Skyfall-GS creates large-scale, high-quality 3D urban scenes using satellite imagery and diffusion models, offering real-time exploration and improved geometry and texture consistency.

Video generation models use the first frame as a conceptual memory buffer, enabling robust customization with minimal training examples.

olmOCR is an open-source toolkit using a fine-tuned vision language model to process PDFs into clean text while preserving structure, optimized for large-scale batch processing.

Qwen3, a unified series of large language models, integrates thinking and non-thinking modes, reduces computational resources, and achieves state-of-the-art performance across various tasks and languages.

Behavior Best-of-N (bBoN) improves the reliability and success rates of computer-use agents by generating and selecting among multiple rollouts using behavior narratives, achieving state-of-the-art performance on OSWorld and strong generalization to different operating systems.

WebWatcher, a multimodal agent with enhanced visual-language reasoning, outperforms existing agents in complex visual and textual information retrieval tasks using synthetic trajectories and reinforcement learning.

A new training framework for RL-based LLM Agents is introduced, extending MDP methodology and demonstrating effectiveness on Multihop QA tasks.