The field of social robot navigation is moving towards more sophisticated and human-aware navigation systems. Researchers are focusing on developing metrics and benchmarks to evaluate the social compliance of robotic agents, as well as learning social heuristics to improve path planning. The use of vision-language models is also being explored to enhance scene understanding and social reasoning in dynamic environments. Notable papers include: Towards Data-Driven Metrics for Social Robot Navigation Benchmarking, which proposes a data-driven approach to benchmarking social robot navigation. Learning Social Heuristics for Human-Aware Path Planning, which introduces a method to learn social value functions for path planning. T-araVLN, which improves vision-and-language navigation in agricultural domains. SocialNav-SUB, which provides a benchmark for evaluating vision-language models in social robot navigation scenarios.