⚠️ Our paper "Neural Conditional Transport Maps" is accepted at @tmlrorg.bsky.social
🧠 We use hypernetworks to learn optimal transport maps that adapt to conditioning variables
👩💻 With @leo-ch.bsky.social , E. Borgonovo and @maxtav.bsky.social
🔗 openreview.net/pdf?id=CZvkp...
Our paper "A dataset of harmonized global air quality monitoring metadata" has been published on Nat. Scientific Data 🛰️
Through transfer learning and geospatial embeddings, we build a large-scale dataset of air quality stations 🫁
🧑🔬 S. Renna, @laraalereis.bsky.social
www.nature.com/articles/s41...
You shall not pass - la version 2025 de Sir Ian McKellen #ICE
New paper: benjaminmoll.com/SRL/
"By preaching that technology shaped by capital is the ultimate horizon, Aghion implicitly accepts that nature, labour and society are second-order variables, subordinated to a fetishized technical devenir". @nobelprize.bsky.social @newleftreview.bsky.social
newleftreview.org/sidecar/post...
Leonardo Chiani, Emanuele Borgonovo, Elmar Plischke, Massimo Tavoni: gsaot: an R package for Optimal Transport-based sensitivity analysis https://arxiv.org/abs/2507.18588 https://arxiv.org/pdf/2507.18588 https://arxiv.org/html/2507.18588
Comments are always more than welcome, hope you find it interesting!
We then evaluate strategies based on their robustness properties. We highlight that a robust policy limits exposure to foreign investors while defending a strong markdown on green debt.
This exercise has also positive implications, as it highlights the key areas of vulnerability faced by Colombia
We first highlight trade-offs across the dimensions of our safe operating space, and iron out three groups of resource mobilisation strategies.
These strategies include the share of green debt (i.e potentially subject to a markdown), the share of foreign exposure within this green debt and the size of the markdown (the "greenium").
To counter this deep uncertainty, policies must be robust, i.e., limit the payoff variance across uncertainty. To get robust strategies, we apply Multi-Objective Robust Decision-Making, and design policies limiting overshoot from a macro safe-operating space on 11 key socio-economic dimensions.
Some developing countries like Colombia face strong external in funding their low-carbon transition. This forces them to design sound resource mobilisation strategies to fund relevant public investment. But thye have to do so in a highly and increasingly uncertain macroeconomic environment.
📢New Working Paper Out!📢
Robust financing mixes for Colombia’s climate strategy - A Robust Decision-Making Application of the GEMMES-Colombia model, with my amazing co-authors #AntoineGodin @celineguivarch.bsky.social and #DevrimYilmaz!
Available here 👇👇
papers.ssrn.com/sol3/papers....
A short 🧵
Net-Zero: A Comparative Study on Neural Network Design for Climate-Economic PDEs Under Uncertainty
https://arxiv.org/abs/2505.13264
