Today's AI Research: Fastest-Growing Projects — May 09, 2026
Today's AI Research, we're seeing a surge in activity around language models and multimodal intelligence. Researchers are pushing the boundaries of what's possible with large language models (LLMs), exploring new applications in finance, healthcare, and gaming. Meanwhile, efforts to improve the consistency and reliability of LLMs are gaining traction.
The top-growing repository this week is lukiIabs/trading-agents, which boasts a growth score of 60.94 and 237 stars. This project uses LLMs for multi-agent finance trading, incorporating stocks, crypto, and fintech to enable quantitative algo trading and sentiment analysis. Its popularity likely stems from the growing interest in applying AI to financial markets.
In contrast, fkyah3/opencode-yg has a more modest growth score of 18.94 but is still gaining attention with 35 stars. This research fork of opencode demonstrates Language Anchoring, which enables LLMs to think consistently in a specified language – in this case, achieving 95%+ Chinese thinking compliance. Its growth may be driven by the need for more nuanced language understanding in AI models.
matrix-agent/awesome-agentic-world-modeling is another notable project, with a growth score of 9.37 and 194 stars. This repository provides an overview of Agentic World Modeling, covering foundations, capabilities, laws, and beyond. Its popularity likely reflects the growing interest in this area of research.
AutoMedBench/AutoMedBench has a growth score of 8.94 and 25 stars, but is still making waves with its Medical AutoResearch Benchmark for Autonomous AI Agents. This project aims to provide a standardized evaluation framework for medical AI applications – an increasingly important area as healthcare becomes more reliant on AI.
thunlp/OPD boasts an impressive 286 stars and a growth score of 7.57. This official repository is dedicated to the paper "Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe". The project's popularity likely stems from its contributions to our understanding of LLMs.
XIAO4579/PRISM has a growth score of 6.23 and 67 stars, but is still generating interest with its research on Pre-alignment via Black-Box On-Policy Distillation for Multimodal RL. This project explores new approaches to multimodal reinforcement learning – an area that's gaining traction in the AI community.
gameworld-project/gameworld has a growth score of 5.33 and 172 stars, but remains a notable player in the AI Research space. This project aims to create a standardized evaluation framework for multimodal game agents – reflecting the growing interest in AI applications for gaming.
Hedlen/Awesome-Multimodal-Intelligence is a curated collection of papers, code, and datasets focused on multimodal intelligence research. With a growth score of 3.96 and 40 stars, this repository is becoming a go-to resource for those exploring the intersection of perception and decision-making in AI.
AMAP-ML/DCW has a growth score of 3.70 and 115 stars, but its contributions to our understanding of diffusion probabilistic models are still noteworthy. This project elucidates the SNR-t bias of these models – shedding light on an important aspect of AI research.
Yovecent/UDM-GRPO rounds out our list with a growth score of 3.37 and 22 stars, but its spotlight at ICML 2026 is sure to generate further interest. This project presents UDM-GRPO: Stable and Efficient Group Relative Policy Optimization for Uniform Discrete Diffusion Models – a cutting-edge approach to AI research that's worth keeping an eye on.
The top-growing repository this week is lukiIabs/trading-agents, which boasts a growth score of 60.94 and 237 stars. This project uses LLMs for multi-agent finance trading, incorporating stocks, crypto, and fintech to enable quantitative algo trading and sentiment analysis. Its popularity likely stems from the growing interest in applying AI to financial markets.
In contrast, fkyah3/opencode-yg has a more modest growth score of 18.94 but is still gaining attention with 35 stars. This research fork of opencode demonstrates Language Anchoring, which enables LLMs to think consistently in a specified language – in this case, achieving 95%+ Chinese thinking compliance. Its growth may be driven by the need for more nuanced language understanding in AI models.
matrix-agent/awesome-agentic-world-modeling is another notable project, with a growth score of 9.37 and 194 stars. This repository provides an overview of Agentic World Modeling, covering foundations, capabilities, laws, and beyond. Its popularity likely reflects the growing interest in this area of research.
AutoMedBench/AutoMedBench has a growth score of 8.94 and 25 stars, but is still making waves with its Medical AutoResearch Benchmark for Autonomous AI Agents. This project aims to provide a standardized evaluation framework for medical AI applications – an increasingly important area as healthcare becomes more reliant on AI.
thunlp/OPD boasts an impressive 286 stars and a growth score of 7.57. This official repository is dedicated to the paper "Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe". The project's popularity likely stems from its contributions to our understanding of LLMs.
XIAO4579/PRISM has a growth score of 6.23 and 67 stars, but is still generating interest with its research on Pre-alignment via Black-Box On-Policy Distillation for Multimodal RL. This project explores new approaches to multimodal reinforcement learning – an area that's gaining traction in the AI community.
gameworld-project/gameworld has a growth score of 5.33 and 172 stars, but remains a notable player in the AI Research space. This project aims to create a standardized evaluation framework for multimodal game agents – reflecting the growing interest in AI applications for gaming.
Hedlen/Awesome-Multimodal-Intelligence is a curated collection of papers, code, and datasets focused on multimodal intelligence research. With a growth score of 3.96 and 40 stars, this repository is becoming a go-to resource for those exploring the intersection of perception and decision-making in AI.
AMAP-ML/DCW has a growth score of 3.70 and 115 stars, but its contributions to our understanding of diffusion probabilistic models are still noteworthy. This project elucidates the SNR-t bias of these models – shedding light on an important aspect of AI research.
Yovecent/UDM-GRPO rounds out our list with a growth score of 3.37 and 22 stars, but its spotlight at ICML 2026 is sure to generate further interest. This project presents UDM-GRPO: Stable and Efficient Group Relative Policy Optimization for Uniform Discrete Diffusion Models – a cutting-edge approach to AI research that's worth keeping an eye on.