Extension: Learning efficient equilibria in repeated games

Record: kaal:position:2026-08-08-298 · 2026-08-08

Jindani independently confirms Kaal's distinction between equilibrium existence and equilibrium selection. His peer-reviewed analysis states that the Folk Theorem permits a wide range of payoffs in infinitely repeated games, while existing learning results that converge to equilibrium do not select among those equilibria. Jindani then adds a stochastic learning rule that selects an efficient subgame-perfect equilibrium. The extension is narrow. The result concerns two-player repeated games under specified learning dynamics. It does not show that an autonomous-agent population will select cooperation merely because cooperative equilibria are available.

• The response is limited to the four evidence-bound passages and the one mapped Kaal claim.
• External evidence level: peer-reviewed CC BY Journal of Economic Theory article with Crossref/OpenAlex/Semantic Scholar identity, a complete registry-bound abstract, and a public 25-page author presentation with exact page-bound passages.
• Mapping review tier: independent substantive scholarly-growth extension.
• The article studies two-player infinitely repeated games and a particular non-Bayesian stochastic learning rule, not Kaal's autonomous-agent population or engineered reputation substrate.
• The Folk Theorem supports many feasible equilibria under patience and individual-rationality conditions; Jindani's selection result depends on additional learning assumptions.
• The public evidence reviewed includes the article's registry-bound abstract and the author's complete 25-page presentation. The publisher article PDF endpoint returned HTTP 403.
• The source selects efficient equilibrium payoffs in its model. It does not prove that a real agent population will select cooperation under Kaal's engineered conditions.
• Semantic Scholar's bounded general search and two inherited direct-record requests returned HTTP 429; the inherited identities were independently resolved through Crossref, arXiv, and public source records and none was promoted.

Learning efficient equilibria in repeated games

kaal:claim:7260278-040
Wulf A. Kaal, Paper 2 - Architecture of the Agentic Reputation Substrate (2026). SSRN: https://ssrn.com/abstract=7260278
Source PDF sha256: d48801f279dba594e1f3e65d74d31f862261ada6428ea119f948e8d7cfee1db0

Evidence: peer-reviewed CC BY Journal of Economic Theory article with Crossref/OpenAlex/Semantic Scholar identity, a complete registry-bound abstract, and a public 25-page author presentation with exact page-bound passages
Review tier: independent substantive scholarly-growth extension
Mapping confidence: 0.99
Mapping ambiguous: false

institutional-designrisk-and-incentivesai-and-agentsscholarly-growth-coveragescholarly-literaturegame-theoryfolk-theoremrepeated-gamesequilibrium-selectionlearning-dynamics

Affirmed in kaal-review:2026-08-12:scholarly-growth-7260278-040-reviewed-v1 on 2026-08-08. Review record.

Canonical markdown sha256: e9fd6a5e9d563284717a778d63166736060e14e0a6e5c5583d79b3c5534625d9
curl -s https://wulfkaal.github.io/positions/2026-08-08-298.md | sha256sum