Test-Time Compute refers to the computational resources allocated during model inference, as distinct from training-time compute. Scaling test-time compute allows models to spend more effort on difficult inputs by generating multiple candidate solutions, performing iterative refinement, or conducting structured search over reasoning paths. Techniques include best-of-N sampling, chain-of-thought prompting, Monte Carlo Tree Search, and adaptive early stopping. Research in this area explores the trade-offs between inference cost, token usage, and output quality across task difficulty.
NEC Laboratories America heads to ACL 2026 in San Diego, California, July 2–7, to present accepted papers spanning knowledge updating and memory control in large language models, task-aware cultural alignment, uncertainty-aware reasoning, and adaptive chain-of-thought optimization, representing some of the most active frontiers in NLP and AI research today.
Princeton Office
4 Independence Way, Suite 200
Princeton, NJ 08540
+1 609-520-1555
San Jose Office
2033 Gateway Place, Suite 200
San Jose, CA 95110
+1 408-863-6007
NEC Laboratories America, Inc. (NEC Labs) is the US-based center for NEC Corporation’s global network of corporate research laboratories. Our diverse research groups collaborate with industry, academia and governments to provide disruptive solutions to complex problems. A leader in the integration of IT and network technologies with more than 100 years of expertise, NEC provides a combination of products and solutions that cross-utilize the company’s experience and global resources to meet the complex and ever-changing needs of its customers.