Researchers developed the Sample Set Aggregator (SSA), a compact language model designed to synthesize multiple candidate solutions generated by a larger base LLM into a single, refined answer. This approach improves over naive majority voting by 8% pass@5 on the MATH dataset, demonstrating robust performance across various mathematical reasoning benchmarks while maintaining computational efficiency comparable to much larger models.
This Resource Letter provides a guide to the literature on fully developed turbulence in fluids. It is restricted to mechanically driven turbulence in an incompressible fluid described by the Navier-Stokes equations of hydrodynamics, and places greatest emphasis on fundamental physical questions. Journal articles and books are cited for the following topics: The Navier-Stokes equations, qualitative aspects of turbulence, the 1941 Kolmogorov theory, intermittency and small scale structure, time correlations and pressure; with brief mention of two-dimensional turbulence, passive scalars in turbulence, and the turbulent boundary layer,
There are no more papers matching your filters at the moment.