6 papers
On the Limits of Prompt-Conditioned Language Models as General-Purpose Learners
David Mguni, Julian Ma, Jun Wang
Large Language Models (LLMs) are frequently portrayed as general-purpose solvers capable of solving arbitrary tasks. We argue that this view overlooks a fundamental constraint: lan…
Multiple Proposer Transaction Fee Mechanism Design: Robust Incentives Against Censorship and Bribery
Aikaterini-Panagiota Stouka, Julian Ma, Thomas Thiery
Transaction Fee Mechanism (TFM) design in blockchain protocols has gained significant attention following the pioneering work of Roughgarden [EC' 21], which established a formal fr…
On the Geometry of Games and their Solvers
Yaqi Sun, Julian Ma, David Mguni
A central challenge in game theory and learning systems such as GANs is understanding which algorithms can efficiently compute equilibria across the heterogeneous landscape of game…
Market Inefficiency in Cryptoasset Markets
Joel Hasbrouck, Julian Ma, Fahad Saleh +1
We demonstrate market inefficiency in cryptoasset markets. Our approach examines investments that share a dominant risk factor but differ in their exposure to a secondary risk. We…
Emergent Bayesian Behaviour and Optimal Cue Combination in LLMs
Julian Ma, Jun Wang, Zafeirios Fountas
Large language models (LLMs) excel at explicit reasoning, but their implicit computational strategies remain underexplored. Decades of psychophysics research show that humans intui…
The Cost of Permissionless Liquidity Provision in Automated Market Makers
Julian Ma, Davide Crapis
Automated market makers (AMMs) allocate fee revenue \textit{proportional} to the amount of liquidity investors deposit. In this paper, we study the economic consequences of the com…