AI research and industry are increasingly focused on autonomous agents, yet definitions of autonomy vary across domains. To understand its evolution, we revisit how autonomy and autonomous agents were defined in research from the 1930s to the 2000s.
This blog post delves into the key research innovations that have shaped each milestone in the evolution from DeepSeekMoE to DeepSeekR1 and DeepSeekV3. It highlights the advancements made in model optimization, reasoning capabilities, and computational efficiency that scale large AI models.
This blog post explores the concepts of backdoor attacks in AI systems, including their sources, categories, and current defense strategies. Notably, backdoor attacks are prevalent in both vision and language-based tasks.
We will explore compound AI systems while highlighting how it can be used to tackle challenges in prompting and real-world problems. We will also compare between Multi-Agent Systems and Compound AI Systems based on shared characteristics.
This annual reflection focuses on the health, knowledge, and skills I have developed.