$R^3$-Bench: LLMs Struggle with Resource-Rational Reasoning under Shared Budgets
In cognitive science, resource rationality asks how an agent should allocate limited computation to maximize expected value. Most reasoning and agent benchmarks use independent per-task budgets; existing shared-budget studies do not calibrate suite performance against the same model's demonstrated single-problem competence. We introduce $R^3$-Bench, which evaluates six-problem suites under shared budgets across mathematics, competitive programming, and abstract reasoning in tool-free and agentic settings. Matched single-problem response curves define an offline empirical oracle over observed successes. Across 72 main-table cells for six models, the oracle mean matches or exceeds the conte...
| Authors | Peisong Wang, Zhiwei Ma, Bowen Liu et al. |
|---|---|
| Primary category | cs.CL |
| Published | 17 August 2026 |
| Last updated at source | 17 August 2026 |
Connected in the knowledge graph
- R-3-Bench/R-3-BenchDataset · cited by
Models and datasets referencing this paper
- R-3-Bench/R-3-BenchDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.