Update README.md

README.md

@@ -19,17 +19,7 @@ Acceleration **O**n **G**PUs) is a CUDA kernel bundle for the structured
 linear-algebra primitives that underpin state-space models (S4/S5, Mamba),
 long-convolution networks (Hyena, H3), and low-displacement-rank layers.
 
-The kernel exposes:
-
-$$K_{b,j} = \sum_i \frac{v_{b,i}}{\omega_j - \lambda_i},\quad
-y = T(c,r)\,x,\quad
-T(c,r)\,x = b,\quad
-u = T^{-1} e_1,\; v = T^{-1} e_n,\quad
-C = F\,T\,F^{-1},$$
-
-$$y = A\,x\;\text{where}\;Z_1 A - A\,Z_{-1} = G H^\top,\;G,H \in \mathbb{R}^{n\times r}$$
-
-— with full autograd, BF16/FP16/complex-dtype paths, stream-aware dispatch,
+The kernel exposes the usual Toeplitz/Cauchy ops — with full autograd, BF16/FP16/complex-dtype paths, stream-aware dispatch,
 and `torch.compile` compliance via `pt2_compliant_tag` +
 `needs_fixed_stride_order`.
 
@@ -184,4 +174,11 @@ nix run -L --max-jobs 1 --cores 4 .#build-and-copy
 - Source: https://github.com/chrisvoncsefalvay/hotdog
 - Hub: https://hf.co/chrisvoncsefalvay/hotdog
 
-Licence: MIT. Author: Chris von Csefalvay.
+## Author
+
+
+I'm [Chris von Csefalvay](chrisvoncsefalvay.com), an AI researcher specialising in post-training, and the author of _[Post-Training: A Practical Guide for
+AI Engineers and Developers](https://posttraining.guide)_ (No Starch Press, 2026). I also write [Post-Slop](https://postslop.substack.com), a periodic diatribe about AI, and what it's doing for society. You can also find me on [LinkedIn](https://linkedin.com/in/chrisvoncsefalvay) and [X](https://x.com/epichrisis).
+## License
+
+MIT