The corresponding repo <a href="https://github.com/VictorTaelin/LamBench" rel="nofollow">https://github.com/VictorTaelin/LamBench</a> describes this as:<p><pre><code> λ-bench A benchmark of 120 pure lambda calculus programming problems for AI models. → Live results What is this? λ-bench evaluates how well AI models can implement algorithms using pure lambda calculus. Each problem asks the model to write a program in Lamb, a minimal lambda calculus language, using λ-encodings of data structures to implement a specific algorithm. The model receives a problem description, data encoding specification, and test cases. It must return a single .lam program that defines @main. The program is then tested against all input/output pairs — if every test passes, the problem is solved. </code></pre> "Live results" wrongly links to <a href="https://victortaelin.github.io/LamBench/" rel="nofollow">https://victortaelin.github.io/LamBench/</a> rather than the correct <a href="https://victortaelin.github.io/lambench/" rel="nofollow">https://victortaelin.github.io/lambench/</a><p>An example task (writing a lambda calculus evaluator) can be seen at <a href="https://github.com/VictorTaelin/lambench/blob/main/tsk/algo_evl.tsk" rel="nofollow">https://github.com/VictorTaelin/lambench/blob/main/tsk/algo_...</a>