EPS
← All batches·2605.16315

A Structural Threshold in Decision Capacity Governs Collapse in Self-Play Reinforcement Learning

topic: current_projecttop score: 100released: 2026-05-19first surfaced: 2026-05-19arXivPDFlinked_to_results2026-05-19

Authors: Arahan Kujur

arXiv · PDF

Summary

arXiv:2605. 16315v1 Announce Type: new Abstract: We show that a threshold in decision capacity determines whether self-play reinforcement learning agents collapse under asymmetric rule perturbations.

Relevance

Read next because A Structural Threshold in Decision Capacity Governs Collapse in Self-Play Reinforcement Learning overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Add C2 control arm (donor sees marker_B without marker_A) to disambiguate paired-marker binding from marker_B leaking alone". Matching terms: under, line, control, full, test. Source: arxiv cs.LG (Machine Learning).

Abstract

arXiv:2605.16315v1 Announce Type: new Abstract: We show that a threshold in decision capacity determines whether self-play reinforcement learning agents collapse under asymmetric rule perturbations. Across poker variants, matrix games, a dice game, and multiple learning algorithms, eliminating all positive-reach contingent decisions causes rapid convergence to a deterministic exploitation attractor, a fixed point at near-maximal loss. Preserving even a single positive-reach contingent decision point prevents this collapse. A frozen baseline and fixed-opponent control confirm that the mechanism is co-adaptation under constraint, not the perturbation itself. The phenomenon is timing-invariant, fully reversible upon action restoration, and intensifies under function approximation. These results establish a sharp threshold at zero reach-weighted contingent action capacity, with severity scaling continuously via reach-weighted capacity in the tested domains.