5 papers
Local Stability and Gaussian Smoothing of Quantized Neural Networks
Sergey Salishev, Anton Makarov, Oleg Granichin
We study Gaussian averaging as a smooth surrogate for quantized neural models. Under bounded local oscillation, we derive a local dimension-dependent bound on |f-g|, linking Gaussi…
Synthesis of signal processing algorithms with constraints on minimal parallelism and memory space
Sergey Salishev
This thesis develops signal-processing algorithms and implementation schemes under constraints of minimal parallelism and memory space, with the goal of improving energy efficiency…
The Narrow Corridor of Stable Solutions in an Extended Osipov--Lanchester Model with Constant Total Population
Sergey Salishev
This paper considers a modification of the classical Osipov--Lanchester model in which the total population of the two forces is preserved over time. It is shown that the d…
GDNSQ: Gradual Differentiable Noise Scale Quantization for Low-bit Neural Networks
Sergey Salishev, Ian Akhremchik
Quantized neural networks can be viewed as a chain of noisy channels, where rounding in each layer reduces capacity as bit-width shrinks; the floating-point (FP) checkpoint sets th…
Regular mixed-radix DFT matrix factorization for in-place FFT accelerators
Sergey Salishev
The generic vector memory based accelerator is considered which supports DIT and DIF FFT with fixed datapath. The regular mixed-radix factorization of the DFT matrix coherent with…