1TradingAgents: Multi-Agents LLM Financial Trading FrameworkA multi-agent framework using large language models for stock trading simulates real-world trading firms, improving performance metrics like cumulative returns and Sharpe ratio.Yijia Xiao2024-12-28140610.8万
2YuE: Scaling Open Foundation Models for Long-Form Music GenerationYuE, a family of open foundation models based on LLaMA2, can generate long-form music with aligned lyrics, coherent structure, and appropriate accompaniment using innovative techniques in next-token pRuibin Yuan2025-03-127739772
3Apodex 1.1: Scaling Agentic Intelligence for Complex WorkApodex 1.1 improves sustained, verifiable progress on complex real-world tasks by scaling executable environments and training agents to coordinate long-horizon work with state maintenance and recoverApodex08-2420734050
4Dream-RSI: Recursive Self-Improvement through Evolving WorldsDream-RSI enables scalable recursive self-improvement by using historical discovery replay to evaluate exploration policies offline, reducing costly online evaluations.Google09-142293821
5A decoder-only foundation model for time-series forecastingA large language model adapted for time-series forecasting achieves near-optimal zero-shot performance on diverse datasets across different time scales and granularities.Abhimanyu Das2023-10-154213.3万
6Paper2Agent: Reimagining Research Papers As Interactive and Reliable AI AgentsPaper2Agent converts research papers into interactive AI agents to facilitate knowledge dissemination and enable complex scientific queries through natural language.taesiri2025-09-094573081
7OpenDevin: An Open Platform for AI Software Developers as Generalist AgentsOpenDevin 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.AK2024-07-248778.9万
8AutoDev: Automated AI-Driven DevelopmentAutoDev is an AI-driven software development framework that automates complex engineering tasks within a secure Docker environment, achieving high performance in code and test generation.Michele Tufano2024-03-132022.5万
9Atria Dawn: The Dawn of Agentic SuperintelligenceAtria Dawn Preview is a foundation agentic language model trained through verified tool interactions that achieves strong benchmark results and demonstrates a shift toward human-AI project-level collaIntern Large Models09-144083503
10SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent HarnessAs coding agents move from supervised code completion to unattended, around-the-clock exploration, their work expands from isolated predictions into long trajectories of reasoning, tool use, and feedbNVIDIA09-176832344
11ZGCM-1: A Fully Open and Extremely Efficient Foundation Model for Math and Agentic SearchZGCM-1 is a 7B open foundation model that combines internal reasoning with external tool use, trained via efficient architecture-system co-design, progressive long-context scaling, and autonomous agenZGCAGI09-112507459
12HuggingFace's Transformers: State-of-the-art Natural Language ProcessingTransformers library provides state-of-the-art Transformer architectures and pretrained models for natural language processing tasks with a unified API and emphasis on extensibility and robust deploymHugging Face2019-10-0929716.6万
13AI for Games in the Foundation Model EraFoundation models, alongside advances in learned game-world models, are reshaping AI across the game lifecycle. Beyond playing games, recent systems model players and game dynamics, support design andNational University of Singapore09-151292198
14RSIAgent: Autonomous Exploration for Recursive Self-improvement in New EnvironmentsRSIAgent is a training-free multi-agent framework that enables recursive self-improvement via autonomous memory construction and broad-then-deep exploration to adapt digital agents to new environmentsAetherLabs-AI09-14752348
15The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing PredictionMixture-of-experts (MoE) inference on consumer hardware is bounded by weight memory: a 35B-class model is 19.5GB at 4-bit, and sparsity shrinks the compute per token, not the bytes that must be held. Edge009-161042002
16FreeToken: Efficient Edge-Native MoE Serving with Bandwidth-Adaptive ExecutionFreeToken is an edge-native Mixture-of-Experts serving system that dynamically maps computation and model state onto heterogeneous local hardware to run large open-weight models on personal machines.University of California, Berkeley08-1710921.3万
17Kronos: A Foundation Model for the Language of Financial MarketsKronos, 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-trainYu Shi2025-08-025443.9万
18Efficient Memory Management for Large Language Model Serving with PagedAttentionPagedAttention algorithm and vLLM system enhance the throughput of large language models by efficiently managing memory and reducing waste in the key-value cache.AK2023-09-126918.6万
19SmolDocling: An ultra-compact vision-language model for end-to-end multi-modal document conversionSmolDocling 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.IBM Granite2025-03-15174196.7万
20MinerU2.5: A Decoupled Vision-Language Model for Efficient High-Resolution Document ParsingMinerU2.5, 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.taesiri2025-09-2617728万
