Today's AI Research: Fastest-Growing Projects — May 05, 2026
Today's AI Research, we're seeing a surge in interest around tools that enable more transparent and explainable AI decision-making. Repositories focused on diagnosing AI misalignment and ensuring language consistency are gaining traction, while research benchmarks for autonomous agents and multimodal intelligence continue to attract attention.
fkyah3/opencode-yg is a research fork of opencode demonstrating Language Anchoring, which makes LLMs think consistently in a specific language, with 95%+ Chinese thinking compliance. With a growth score of 24.65 and 32 stars, this repository is growing rapidly as researchers seek to improve the reliability of large language models.
ifixai-ai/diagnostic offers an open-source diagnostic for AI misalignment, providing 32 tests across fabrication, manipulation, deception, unpredictability, and opacity. Its provider-agnostic approach and ability to run against multiple AI platforms have earned it a growth score of 18.06 and 160 stars, as researchers and developers seek to ensure the integrity of their AI systems.
AutoMedBench/AutoMedBench is a Medical AutoResearch Benchmark for Autonomous AI Agents, providing a framework for evaluating the performance of autonomous agents in medical settings. With a growth score of 11.65 and 24 stars, this repository is growing as researchers explore the potential applications of autonomous AI in healthcare.
matrix-agent/awesome-agentic-world-modeling offers a curated collection of resources on Agentic World Modeling, covering foundations, capabilities, laws, and beyond. Its comprehensive approach has earned it a growth score of 11.59 and 183 stars, as researchers delve into the complexities of agentic world modeling.
thunlp/OPD is the official repository for the paper "Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe". With a growth score of 7.57 and 225 stars, this repository is growing as researchers investigate new approaches to on-policy distillation.
7WaySecurity/ai_osint provides curated AI OSINT resources, including Google dorks, Shodan queries, and techniques for discovering exposed LLM endpoints and leaked AI API keys. Its growth score of 6.63 and 76 stars reflect the increasing importance of security in AI research.
gameworld-project/gameworld is a standardized evaluation framework for multimodal game agents, enabling verifiable assessment of agent performance. With a growth score of 6.30 and 168 stars, this repository is growing as researchers explore the potential applications of multimodal game agents.
zubair-trabzada/ai-trading-claude is an AI trading research engine that analyzes stocks, options strategies, and sector rotation using technical, fundamental, sentiment, risk, and thesis analysis. Its growth score of 5.50 and 101 stars reflect the growing interest in AI-driven financial analysis.
Hedlen/Awesome-Multimodal-Intelligence offers a curated collection of resources on multimodal intelligence research, covering VLMs, VLAs, world models, and embodied AI. With a growth score of 5.33 and 36 stars, this repository is growing as researchers investigate the potential applications of multimodal intelligence.
lukiiabs/trading-agents provides an LLM multi-agent finance trading platform that enables quantitative algo trading, sentiment analysis, and more. Its growth score of 4.86 and 233 stars reflect the increasing interest in AI-driven trading platforms.
fkyah3/opencode-yg is a research fork of opencode demonstrating Language Anchoring, which makes LLMs think consistently in a specific language, with 95%+ Chinese thinking compliance. With a growth score of 24.65 and 32 stars, this repository is growing rapidly as researchers seek to improve the reliability of large language models.
ifixai-ai/diagnostic offers an open-source diagnostic for AI misalignment, providing 32 tests across fabrication, manipulation, deception, unpredictability, and opacity. Its provider-agnostic approach and ability to run against multiple AI platforms have earned it a growth score of 18.06 and 160 stars, as researchers and developers seek to ensure the integrity of their AI systems.
AutoMedBench/AutoMedBench is a Medical AutoResearch Benchmark for Autonomous AI Agents, providing a framework for evaluating the performance of autonomous agents in medical settings. With a growth score of 11.65 and 24 stars, this repository is growing as researchers explore the potential applications of autonomous AI in healthcare.
matrix-agent/awesome-agentic-world-modeling offers a curated collection of resources on Agentic World Modeling, covering foundations, capabilities, laws, and beyond. Its comprehensive approach has earned it a growth score of 11.59 and 183 stars, as researchers delve into the complexities of agentic world modeling.
thunlp/OPD is the official repository for the paper "Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe". With a growth score of 7.57 and 225 stars, this repository is growing as researchers investigate new approaches to on-policy distillation.
7WaySecurity/ai_osint provides curated AI OSINT resources, including Google dorks, Shodan queries, and techniques for discovering exposed LLM endpoints and leaked AI API keys. Its growth score of 6.63 and 76 stars reflect the increasing importance of security in AI research.
gameworld-project/gameworld is a standardized evaluation framework for multimodal game agents, enabling verifiable assessment of agent performance. With a growth score of 6.30 and 168 stars, this repository is growing as researchers explore the potential applications of multimodal game agents.
zubair-trabzada/ai-trading-claude is an AI trading research engine that analyzes stocks, options strategies, and sector rotation using technical, fundamental, sentiment, risk, and thesis analysis. Its growth score of 5.50 and 101 stars reflect the growing interest in AI-driven financial analysis.
Hedlen/Awesome-Multimodal-Intelligence offers a curated collection of resources on multimodal intelligence research, covering VLMs, VLAs, world models, and embodied AI. With a growth score of 5.33 and 36 stars, this repository is growing as researchers investigate the potential applications of multimodal intelligence.
lukiiabs/trading-agents provides an LLM multi-agent finance trading platform that enables quantitative algo trading, sentiment analysis, and more. Its growth score of 4.86 and 233 stars reflect the increasing interest in AI-driven trading platforms.