On The Statistical Limits Of Self-improving Agents

Abstract

As systems trend toward superintelligence, a natural modeling premise is that agents can self-improve along every facet of their own design. We formalize this with a five-axis decomposition and a decision layer, separating incentives from learning behavior and analyzing axes in isolation. Our central result identifies and introduces a sharp utility-learning tension, the structural conflict in self-modifying systems whereby utility-driven changes that improve immediate or expected performance can also erode the statistical preconditions for reliable learning and generalization. Our findings show that distribution-free guarantees are preserved iff the policy-reachable model family is uniformly capacity-bounded; when capacity can grow without limit, utility-rational self-changes can render learnable tasks unlearnable. Under standard assumptions common in practice, these axes reduce to the same capacity criterion, yielding a single boundary for safe self-modification.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keywang2025on

Related papers

• A Mathematical Theory Of Agency And Intelligence (2026)0.00
• Winning Isn't Everything: Enhancing Game Development With Intelligent Agents (2019)11.29
• Embedded Universal Predictive Intelligence: A Coherent Framework For Multi-agent Learning (2025)0.00
• On The Limited Representational Power Of Value Functions And Its Links To Statistical (in)efficiency (2024)0.00
• Modulation Of Viability Signals For Self-regulatory Control (2020)0.00
• Impact Of Decentralized Learning On Player Utilities In Stackelberg Games (2024)0.00
• Reasoning About Unforeseen Possibilities During Policy Learning (2018)0.00
• Dynamics Of Moral Behavior In Heterogeneous Populations Of Learning Agents (2024)4.52