Different transformers may have very different weights, but they often hide the same tiny "engine" inside that actually does the task.
The paper finds a simple trick—randomly skipping some parameter updates—can train large language models better than fancy optimizers.
This survey asks how close AI memory systems are to human memory and organizes the answer into three parts: implicit memory (inside the model), explicit memory (outside storage you can look up), and agentic memory (what an AI agent keeps over time to plan and act).