Trending Papers

SkillOpt introduces a systematic text-space optimizer for agent skills that trains skills as external agent state with stable updates and zero deployment inference overhead, achieving superior performance across multiple benchmarks and execution environments.

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.

Kronos, a specialized pre-training framework for financial K-line data, outperforms existing models in forecasting and synthetic data generation through a unique tokenizer and autoregressive pre-training on a large dataset.

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.

EverMemOS presents a self-organizing memory system for large language models that processes dialogue streams into structured memory cells and scenes to enhance long-term interaction capabilities.

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.

Vision Language Models can be adapted for 3D understanding tasks through simple architectural modifications and text-based training, achieving performance comparable to specialized vision models without requiring complex designs or extensive data augmentation.

Stable-worldmodel provides a modular and standardized research framework for developing and evaluating world models with controllable environmental factors for robustness and continual learning applications.

MOSS-TTS is a speech generation model using discrete audio tokens and autoregressive modeling with capabilities for voice cloning, pronunciation control, and long-form generation across multiple languages.

A comprehensive framework is presented for converting bidirectional video diffusion models into real-time interactive world models with controllable, causal, and low-latency capabilities through fine-tuning and distillation techniques.

Mixture of vision encoders and resolutions in multimodal large language models improves performance through concatenation of visual tokens and a Pre-Alignment mechanism, leading to superior results on benchmarks.

OpenDevin is a platform for developing AI agents that interact with the world by writing code, using command lines, and browsing the web, with support for multiple agents and evaluation benchmarks.

LLM-based agents perform poorly on VibeSearch benchmark, which evaluates multi-turn dialogue search scenarios reflecting real user-agent collaboration rather than traditional single-turn query tasks.

PaddleOCR-VL, a vision-language model combining NaViT-style dynamic resolution and ERNIE, achieves state-of-the-art performance in document parsing and element recognition with high efficiency.

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.

PagedAttention algorithm and vLLM system enhance the throughput of large language models by efficiently managing memory and reducing waste in the key-value cache.

Person-grounded AI skills are automatically distilled from heterogeneous traces into inspectable, correctable packages that capture both capabilities and behavioral patterns.

ARIS is an open-source research harness that uses cross-model adversarial collaboration to ensure reliable long-term research outcomes through coordinated execution, orchestration, and assurance layers.

SmolDocling is a compact vision-language model that performs end-to-end document conversion with robust performance across various document types using 256M parameters and a new markup format.

SANA-Video, a small diffusion model, efficiently generates high-resolution, high-quality videos with strong text-video alignment using linear attention and a constant-memory KV cache, achieving competitive performance at a lower cost and faster speed.

MobileGym presents a browser-based mobile environment enabling deterministic evaluation and scalable reinforcement learning through JSON-based state management and parallel execution.

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.

Enhancements to the AgentScope platform improve scalability, efficiency, and ease of use for large-scale multi-agent simulations through distributed mechanisms, flexible environments, and user-friendly tools.

AI-Trader presents the first fully automated live benchmark for evaluating large language models in financial decision-making across multiple markets with autonomous information processing.

Generative multimodal foundation models are used to create high-quality training data for image restoration, improving model generalization across diverse real-world scenarios.

SkillNet introduces an open infrastructure for systematically accumulating and transferring AI skills through a unified ontology, significantly improving agent performance across multiple domains.

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.

AutoResearchClaw is a multi-agent autonomous research system that improves scientific discovery through structured debate, self-healing execution, verifiable reporting, human collaboration, and evolutionary learning, outperforming previous systems on a benchmark while maintaining human oversight.

Mega-ASR framework improves robustness in real-world speech recognition through compound-data construction and progressive acoustic-to-semantic optimization techniques.

TriSplat is a feed-forward 3D reconstruction network that uses oriented triangle primitives to directly generate simulation-ready meshes from single images, bypassing expensive post-processing steps.

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.

GenericAgent is a self-evolving large language model agent system that maximizes context information density through hierarchical memory, reusable SOPs, and efficient compression to overcome long-horizon limitations.

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

PDFMathTranslate enables layout-preserving scientific document translation using large language models and precise layout detection, offering improved precision, flexibility, and efficiency.

MiniCPM4, a highly efficient large language model for end-side devices, achieves superior performance using innovations in sparse attention, pre-training datasets, training algorithms, and inference systems.

A large language model adapted for time-series forecasting achieves near-optimal zero-shot performance on diverse datasets across different time scales and granularities.

GenClaw presents a code-driven agentic image generation framework that enables precise visual construction through conceptualization, sketching, and coloring stages, integrating programmatic logic with generative models.

DataFlow is an LLM-driven data preparation framework that enhances data quality and reproducibility for various tasks, improving LLM performance with automatically generated pipelines.

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.

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.

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 Hierarchical Recurrent Model architecture with specialized training on instruction-response pairs achieves competitive language modeling performance with significantly reduced computational requirements compared to traditional Transformer-based approaches.

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.

Lance is a unified multimodal model that combines understanding, generation, and editing capabilities for images and videos through collaborative multi-task training and a dual-stream architecture.

LongTraceRL addresses long-context reasoning challenges in large language models through tiered distractor construction and rubric reward design for improved reasoning quality.

MiniCPM-o 4.5 enables real-time full-duplex multimodal interaction through Omni-Flow, a unified streaming framework that aligns inputs and outputs temporally for simultaneous perception and response.

MiniCPM-V 4.5, a 8B parameter multimodal large language model, achieves high performance and efficiency through a unified 3D-Resampler architecture, a unified learning paradigm, and a hybrid reinforcement learning strategy.

LongLive-2.0 presents an NVFP4-based parallel infrastructure for long video generation that addresses training and inference bottlenecks through sequence-parallel autoregressive training and diffusion model tuning.

OSCAR is an ultra-low-bit KV cache quantization method that aligns quantization with attention-aware covariance structures, achieving high accuracy and efficiency for long-context LLM serving.

Agentic coding tools receive goals written in natural language as input, break them down into specific tasks, and write or execute the actual code with minimal human intervention. Central to this process are agent context files ("READMEs for agents") that provide persistent, project-level instructions. In this paper, we conduct the first large-scale empirical study of 2,303 agent context files from 1,925 repositories to characterize their structure, maintenance, and content. We find that these files are not static documentation but complex, difficult-to-read artifacts that evolve like configuration code, maintained through frequent, small additions. Our content analysis of 16 instruction types shows that developers prioritize functional context, such as build and run commands (62.3%), implementation details (69.9%), and architecture (67.7%). We also identify a significant gap: non-functional requirements like security (14.5%) and performance (14.5%) are rarely specified. These findings indicate that while developers use context files to make agents functional, they provide few guardrails to ensure that agent-written code is secure or performant, highlighting the need for improved tooling and practices.