update

README.md

@@ -21,7 +21,7 @@ pip install hypergrad
 
 * conjugate gradient
 * [Neumann-series approximation](https://proceedings.mlr.press/v108/lorraine20a.html)
-* [Nyström method](to_be_updated)
+* [Nyström method](https://arxiv.org/abs/2302.09726)
 
 Implementation of these methods can be found in `hypergrad/approximate_ihvp.py`
 
@@ -45,6 +45,6 @@ To cite this repository,
  author = {Ryuichiro Hataya and Makoto Yamada},
  title = {{Nystr\"om Method for Accurate and Scalable Implicit Differentiation}},
  booktitle = {AISTATS},
- year = {2023}
+ year = {2023},
 }
 ```

docs/index.md

@@ -33,7 +33,7 @@ The essential challenge of such nested problems is that $\nabla_\phi g$ needs ba
 f(\theta, \phi; \mathcal{T})$.
 One way to address this is to use *approximated implicit differentiation* methods.
 `hypergrad` currently supports [conjugate gradient-based approximation](), [Neumann-series approximation](),
-and [Nyström method approximation]() like:
+and [Nyström method approximation](https://arxiv.org/abs/2302.09726) like:
 
 ```python
 from hypergrad.approx_hypergrad import nystrom