


PhD Student @UniHannover ERC project (2023-28) on MODEL TRANSFER in Science (#ERCStGr @ERC_research)


My article, "Model Transfer, the Challenge of Tractability Standards, and Methodological Embedding," has been published in Philosophy of Science as FirstView! You can read it open access via the link below.

Where computer scientists and economists talk to each other Search ad auctions, matching markets that place students and organs, pricing and recommendation systems that participants learn to game: each is at once an economic question about how people respond to incentives and a computational question about what can be solved at the scale the market actually runs on. EconCS is what results when both questions are answered together rather than in sequence. Cornell is where much of that conversation began, and this article is the clearest short account of how it happened. It follows the field from informal exchanges between economists and computer scientists in the late 1990s, through the 2005 “Getting Connected” initiative, to the Networks course David Easley and Jon Kleinberg built and the textbook that grew out of it. Readers new to this work should read it first: it conveys what the field is and why it exists in plain language, before any technical paper asks for a formal model. as.cornell.edu/news/where-co…

Game theory" is arguably a misnomer as a single term. The four fields that study it actually share almost entirely different mathematical toolkits .microeconomic theory, TCS, OR, and finance all do game theory. Being able to follow a proof in one branch essentially doesn't mean you can follow one in another. The technical split is fairly complete: Econ theory guys: the envelope theorem, monotonicity, convex analysis, Bayesian persuasion, lattice theory and monotone comparative statics (Topkis, Milgrom–Shannon). Fixed-point theorems get cited, but not used. TCS/AGT: combinatorics, LP duality (the Cai–Devanur–Weinberg line of work), approximation ratios, complexity classes, no-regret online learning, the smoothness framework. OR: variational inequalities and complementarity problems (Nash equilibrium written as a VI, Wardrop traffic equilibrium, Lemke–Howson as an LCP), large-scale numerical computation. Finance: stochastic analysis, martingales, filtrations, HJB, Kyle model.
Kakutani/Brouwer fixed point theorems are not useful in applied game theory. In game theory we write down a game and try to solve for equilibria with nice properties (pure strategy, linearity, etc.) you can just skip the step of proving existence using a fixed point theorem

Мой аккаунт тоже заблочили :(
Мы получаем большое количество обращений по блокировкам в Scalable Capital Если у вас есть какая-то информация помимо «из-за санкций» – пожалуйста сообщите. Мы знаем что брокер ссылается на санкции без конкретики Если у вас есть конкретика – дайте знать, нам это очень поможет в возможном разрешении ситуации

I use this response to illustrate what I call methodological embedding, a model-reformulation strategy for overcoming the challenges posed by differing tractability standards.

One response to this problem by proponents of agent-based modeling has been the construction of "hybrid" models that combine elements of agent-based models with macroeconomic general equilibrium models.

Specifically, I explain the difficulties surrounding the adoption of agent-based models in macroeconomics as a problem of establishing and negotiating tractability.

The existence of varying tractability standards across scientific domains has important implications for #ModelTransfer: a model considered tractable in one context may no longer be regarded as tractable when transferred to another.

In this paper, I challenge the default view of tractability in the philosophy of science, drawing on Béatrice Cherrier's ideas regarding tractability standards.

I am happy to announce that my paper, "Model Transfer, the Challenge of Tractability Standards and Methodological Embedding," has been accepted for publication in Philosophy of Science. You can read the preprint here: philsci-archive.pitt.edu/304…

Can AI "learn" economic states, addressing the Lucas Critique? With @alexolegimas we simulated data from an NK model, fit a transformer, and tested out of sample fit It generalizes surprisingly well. We hope this stimulates discussion and future agendas arpitrage.substack.com/p/can…


My understanding of the "Moll Critique" of rational expectations (RE) with heterogeneous agents (HA). With rep-agent, RE=we look for the key under the lamp-post. Luckily, the key really is there: the simplification both makes the problem tractable and gives the right answer. 1/n

I have just posted my survey paper “Deep Learning for Solving Economic Models” on my webpage: sas.upenn.edu/~jesusfv/Deep_… In one or two weeks, it will also circulate as a working paper at the NBER and CEPR. Still, I wanted to let people know already, since I am quite happy with the outcome, largely thanks to some fantastic early feedback I got. As I have often argued, the ongoing revolution in deep learning is transforming how we solve dynamic equilibrium economic models. At its core, solving a model amounts to approximating unknown target functions (such as the value function of agents, a decision rule, or a best response function). Deep learning frequently does a fantastic job at that task. In the paper, I emphasize that this success is not “magic,” but rather the direct consequence of deep learning’s ability to discover better representations of the relevant variables of a model (for example, the state variables). The layers of a neural network transform the input variables into informationally efficient representations that can be more easily approximated. Tom Sargent loves to say that finding the state is an art. Deep learning tries to automatize that art as much as possible. This is why, in many cases, we can now solve high-dimensional problems that were computationally infeasible only a few years ago. Furthermore, the structure of deep networks designed for solving these models, largely linear apart from the non-linearity encapsulated in the activation function, permits massive parallelization. The survey paper is designed to start from the ground up. My intended audience is a first-year graduate student with only a very basic knowledge of solution methods, or even a motivated senior undergraduate. I would very much appreciate feedback. Can you follow the arguments throughout? Are there steps that remain unclear? I have taught courses based on this material at Penn, the Bank of Spain, Cambridge, the ECB, Harvard, Johns Hopkins, Northwestern, Oxford, Princeton, UC Santa Barbara, and Stanford, but I am always looking for fresh eyes to suggest improvements. All the slide decks, with links to the code, are available here: sas.upenn.edu/~jesusfv/teach… under “Machine Learning for Economists.” Eventually, I may use this survey paper and the slide decks as the kernel for something longer, but first, I need to clear my desk of too many ongoing projects.

Москва 2025 — сон сумасшедшего, ты идешь по центру там ягодные сезоны, за ними вакансии оператора БПЛА, идут куклы на ходулях, рядом — вечер военной поэзии, экспозиция СЕМЬЯ СКВОЗЬ ВЕКА, огромный самовар (???), следом выставка эпизоды финской русофобии