arXiv:2606. 27397v1 Announce Type: cross Abstract: Evaluating LLM agents requires dynamic environments that go beyond static reasoning and zero-sum games.
Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.
Evaluating LLM agents requires dynamic environments that go beyond static reasoning and zero-sum games. Real-world economic interaction is often open-ended and mixed-motive: agents must negotiate, create positive-sum surplus, compete for scarce assets, and plan under delayed returns.
arXiv:2605. 13909v2 Announce Type: replace-cross Abstract: Negotiation is a central mechanism of economic exchange, shaping markets, procurement, labor agreements, and resource allocation.
arXiv:2606. 28943v1 Announce Type: cross Abstract: Learning to bid in repeated multi-unit auctions with bandit feedback poses a fundamental challenge.
arXiv:2607. 05863v1 Announce Type: new Abstract: Negotiation is a fundamental strategic interaction in management science, characterized by agents attempting to reach agreements while protecting private information, such as reservation costs and hidden valuations.
arXiv:2606. 06480v1 Announce Type: cross Abstract: Many real-world competitive systems require multiple decision-makers to act simultaneously under shared constraints, limited information, and repeated interaction, as in auctions, resource allocation, and security competition.
arXiv:2606. 11284v1 Announce Type: cross Abstract: Real-world multi-agent systems, from traffic coordination to resource allocation, are often modeled as general-sum games where individual incentives conflict with collective welfare.