Google, IBM & Meta Certificates – 40% Off
One plan covers every Professional Certificate on Coursera.
Unlock All Certificates
Explore a comprehensive seminar that introduces Domain-Adaptive Programming, a paradigm designed to enhance Large Language Models' performance in specialized, high-stakes domains by integrating formal, symbolic, and programmatic structures. Learn how this approach addresses the brittleness of LLMs in critical applications where reasoning errors and unverified plans can lead to serious failures. Discover methods that combine the flexibility of neural language models with the robustness and verifiability of symbolic systems through structured program generation, latent optimization for multi-step reasoning and planning, and iterative feedback loops with symbolic verification. Examine practical applications across text processing, vision-language tasks, and embodied AI domains, understanding how grounding neural reasoning in formal structures can expand the boundaries of what LLMs can reliably solve. Gain insights from cutting-edge research that bridges natural language processing, machine learning, and vision-language reasoning through neuro-symbolic methodologies, presented by a USC Ph.D. candidate whose work has been published at premier venues including NeurIPS, ICLR, NAACL, EMNLP, CVPR, and ACL.
