# Learning to Manipulate under Limited Information

arXiv: https://arxiv.org/abs/2401.16412 · HTML: https://mlsanity.com/paper/2401.16412/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 57 references that are in the map.
- First posted on arXiv 29 Jan 2024.

## Builds on (most cited first, 8 of 8)

- [Understanding deep learning requires rethinking generalization](https://mlsanity.com/paper/1611.03530/) (2016)
- [Deep Learning Scaling is Predictable, Empirically](https://mlsanity.com/paper/1712.00409/) (2017)
- [Are wider nets better given the same number of parameters?](https://mlsanity.com/paper/2010.14495/) (2020)
- [Learning to Elect](https://mlsanity.com/paper/2108.02768/) (2021)
- [Split Cycle: A New Condorcet Consistent Voting Method Independent of Clones and Immune to Spoilers](https://mlsanity.com/paper/2004.02350/) (2020)
- [Practical Algorithms for Multi-Stage Voting Rules with Parallel Universes Tiebreaking](https://mlsanity.com/paper/1901.09791/) (2019)
- [A Local-Dominance Theory of Voting Equilibria](https://mlsanity.com/paper/1404.4688/) (2014)
- [Modeling Peoples Voting Behavior with Poll Information](https://mlsanity.com/paper/1909.10492/) (2019)

## Built on by (most cited first, 3 of 3)

- [Learning How to Vote with Principles: Axiomatic Insights Into the Collective Decisions of Neural Networks](https://mlsanity.com/paper/2410.16170/) (2024)
- [Learning Resilient Elections with Adversarial GNNs](https://mlsanity.com/paper/2601.01653/) (2026)
- [Can AI mediation improve democratic deliberation?](https://mlsanity.com/paper/2601.05904/) (2026)

## Abstract (arXiv)

By classic results in social choice theory, any reasonable preferential voting method sometimes gives individuals an incentive to report an insincere preference. The extent to which different voting methods are more or less resistant to such strategic manipulation has become a key consideration for comparing voting methods. Here we measure resistance to manipulation by whether neural networks of various sizes can learn to profitably manipulate a given voting method in expectation, given different types of limited information about how other voters will vote. We trained over 100,000 neural networks of 26 sizes to manipulate against 8 different voting methods, under 6 types of limited information, in committee-sized elections with 5-21 voters and 3-6 candidates. We find that some voting methods, such as Borda, are highly manipulable by networks with limited information, while others, such as Instant Runoff, are not, despite being quite profitably manipulated by an ideal manipulator with full information. For the three probability models for elections that we use, the overall least manipulable of the 8 methods we study are Condorcet methods, namely Minimax and Split Cycle.