The field of Edge AI and Large Language Model (LLM) inference is rapidly evolving, driven by the need for efficient, low-latency, and private processing of complex AI workloads. A common theme among recent developments is the focus on improving performance and scalability while reducing computational requirements. Notable advancements include the use of heterogeneous AI extensions, dynamic activation-aware weight pruning, and bandwidth management to enhance multi-core CPU performance. Novel serverless data planes and federated pruning frameworks have also been proposed to improve LLM inference efficiency and scalability. Collaborative inference systems, which leverage both edge and cloud resources, aim to balance latency, cost, and privacy concerns. Some papers are particularly noteworthy, such as EdgeMM, which presents a multi-core CPU solution with heterogeneous AI extensions, achieving a 2.84x performance speedup compared to laptop GPUs. Palladium is also notable, as it introduces a DPU-enabled serverless data plane that reduces CPU burden and enables efficient zero-copy communication, resulting in a 20.9x improvement in RPS and a 21x reduction in latency. In the area of LLMs, researchers are exploring quantization techniques to reduce model precision while maintaining performance. Methods such as pseudo-quantization training, quantization-aware training, and native low-precision training have shown promising results. Innovative methods like Gaussian weight sampling, outlier token tracing, and quartet-based training have achieved scalable, efficient, and stable training. Noteworthy papers include Accurate KV Cache Quantization with Outlier Tokens Tracing and Quartet: Native FP4 Training Can Be Optimal for Large Language Models. The field of Mixture-of-Experts (MoE) is also experiencing significant developments, driven by innovations in expert selection, routing policies, and model compression. Researchers are exploring new methods to enhance MoE model efficiency and effectiveness, including hierarchical task-guided and context-responsive routing policies. Noteworthy papers in this area include THOR-MoE and Efficient Data Driven Mixture-of-Expert Extraction from Trained Networks. Overall, the field is advancing towards more efficient and deployable LLMs through the development of effective quantization techniques, dynamic workload reduction schemes, and efficient model editing methods. Noteworthy papers include TokenWeave, MegaScale-MoE, EfficientLLM, ULTRAEDIT, Polar Sparsity, and LyapLock. These advancements have the potential to significantly improve the performance, scalability, and applicability of Edge AI and LLMs across various applications.