Benchmark details

[NitayRabi]

Hey! awesome work on RLM, I've been working on porting it to Typescript (https://github.com/code-rabi/rllm) with V8 isolation and schematic context support.

I've wanted to know if you can provide a bit more details on the benchmark (dataset, iterations per benchmark par context size)? I tried to run some benchmarking but wanted to get an identical benchmark to see the porting is successful

[alexzhang13]

Hey! The repo containing benchmarks / experiments used in the paper currently aren't out because I'm preparing it for ICML. All details about what was run / how are available in the paper (https://arxiv.org/abs/2512.24601v1) but if you have specific questions, happy to try and answer.