Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
Binary Quantized Neural Network Training Is W[1]-Hard Parameterized by Input and Output Dimensions
Tao Jiang, Minbo Gao, Shaowei Cai
Ganian et al. (ICLR 2026) proved that quantized neural network training is fixed-parameter tractable when parameterized jointly by architecture treewidth, input dimension , and…
cs.LG2026
Exact Minimax One-Bit Unbiased Compression: Heavy-Tail Necessity and Finite-Randomness Approximation
Tao Jiang, Minbo Gao, Shaowei Cai
A pointwise-unbiased one-bit compressor reconstructs every real input in expectation while transmitting one bit. For a scalar source with CDF , mean , and $\mathcal J(P)=…