Trending Papers

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.

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.

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

Kimi Linear, a hybrid linear attention architecture, outperforms full attention in various scenarios with improved efficiency and performance, using Kimi Delta Attention and Multi-Head Latent Attention.

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.

Continuous Autoregressive Language Models (CALM) improve language model efficiency by predicting continuous vectors instead of discrete tokens, reducing computational cost while maintaining performance.

Emu3.5, a large-scale multimodal world model, predicts next states in vision and language, enhanced with reinforcement learning and Discrete Diffusion Adaptation for efficient inference, achieving strong performance in various multimodal tasks.

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.

VCode introduces a benchmark for generating SVG code from images to preserve symbolic meaning, highlighting gaps in visual-centric coding and proposing VCoder to improve performance.

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.

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.

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.

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

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.

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.

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.

The paper proposes a framework for building end-to-end agentic information seeking agents through a combination of data construction, trajectory sampling, supervised fine-tuning, and reinforcement learning, showcasing its effectiveness on information seeking benchmarks.

WebShaper, a formalization-driven framework, synthesizes information-seeking datasets using set theory and Knowledge Projections to enhance reasoning structure and achieve top performance in open-sourced benchmarks.

AgentFounder, a deep research agent model incorporating Agentic Continual Pre-training, achieves state-of-the-art performance in agentic tasks while maintaining strong tool-use ability.

WebWeaver, a dual-agent framework, addresses open-ended deep research challenges by integrating adaptive planning and focused synthesis to produce high-quality, reliable reports.

WebSailor, a post-training methodology, enhances open-source models with systematic uncertainty reduction, matching proprietary agents' performance in complex information-seeking tasks.

ReSum, a novel paradigm with periodic context summarization, enhances web agents' performance on knowledge-intensive tasks by overcoming context window limitations, achieving significant improvements over ReAct.

WebSailor, a post-training methodology, enhances open-source LLMs with sophisticated reasoning to match proprietary systems in complex information-seeking tasks.

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.

Denario, an AI multi-agent system, performs various scientific research tasks and generates papers across multiple disciplines, demonstrating its capabilities and limitations through expert evaluations.

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.

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.

MedRAX, a versatile AI agent integrating advanced CXR analysis tools and multimodal large language models, achieves top performance in a wide range of medical queries without extra training.

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.

Agent S2, a compositional framework using Mixture-of-Grounding and Proactive Hierarchical Planning, achieves state-of-the-art performance in computer use automation across various benchmarks and operating systems.

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.

Diffusion language models outperform autoregressive models in low-data settings due to any-order modeling, iterative bidirectional denoising, and Monte Carlo augmentation, and maintain advantages even at scale.

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.

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.

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.

DeepAgent, an end-to-end deep reasoning agent, autonomously performs thinking, tool discovery, and action execution using memory folding and reinforcement learning, outperforming baselines in various tool-use and application tasks.

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.

LongCat-Video, a 13.6B parameter video generation model based on the Diffusion Transformer framework, excels in efficient and high-quality long video generation across multiple tasks using unified architecture, coarse-to-fine generation, and block sparse attention.

Pico-Banana-400K is a large-scale, high-quality dataset for instruction-based image editing, featuring diverse edit pairs, multi-turn editing, preference subsets, and long-short instruction pairs, enabling comprehensive research and benchmarking.

Cache-to-Cache (C2C) enables direct semantic communication between LLMs using neural network projections, improving accuracy and reducing latency compared to text-based communication.

RLinf-VLA is a unified framework for scalable reinforcement learning training of vision-language-action models, offering improved performance and generalization compared to supervised fine-tuning.

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

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.

ThinkMorph, a unified model fine-tuned on interleaved reasoning traces, enhances multimodal reasoning by generating coherent text-image steps, achieving significant performance gains and demonstrating emergent capabilities.

Dynamic Position Extrapolation (DyPE) enhances ultra-high-resolution image generation by dynamically adjusting positional encodings in pre-trained diffusion transformers, achieving state-of-the-art fidelity without additional sampling cost.

PaperCoder is a multi-agent LLM framework that converts machine learning papers into functional code repositories through planning, analysis, and generation stages.

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

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.

Large Market Model (LMM) and MarS simulate realistic financial market interactions, addressing scalability and realism in financial applications.