feat: mask compute to lazy mask execution

[joseph-isaacs]

Does this PR closes an open issue or discussion?

• Closes #.

What changes are included in this PR?

What is the rationale for this change?

How is this change tested?

Are there any user-facing changes?

[codspeed-hq]

Merging this PR will degrade performance by 49.83%

⚡ 1 improved benchmark
❌ 16 regressed benchmarks
✅ 1118 untouched benchmarks
⏩ 1268 skipped benchmarks¹

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	bench_dict_mask[(0.9, 0.9)]	1.3 ms	2.4 ms	-48.65%
❌	Simulation	bench_dict_mask[(0.01, 0.9)]	1.2 ms	2.4 ms	-49.83%
❌	Simulation	bench_dict_mask[(0.1, 0.9)]	1.2 ms	2.3 ms	-47.32%
❌	Simulation	bench_dict_mask[(0.1, 0.01)]	1.2 ms	2 ms	-38.3%
❌	Simulation	bench_dict_mask[(0.01, 0.1)]	1.2 ms	2 ms	-39.85%
❌	Simulation	bench_dict_mask[(0.5, 0.01)]	1.2 ms	1.8 ms	-34.56%
❌	Simulation	bench_dict_mask[(0.01, 0.5)]	1.2 ms	2.1 ms	-42.34%
❌	Simulation	bench_dict_mask[(0.9, 0.01)]	1.2 ms	2 ms	-38.33%
❌	Simulation	bench_dict_mask[(0.5, 0.1)]	1.2 ms	1.9 ms	-36.21%
❌	Simulation	bench_dict_mask[(0.5, 0.9)]	1.2 ms	2.3 ms	-47.29%
❌	Simulation	bench_dict_mask[(0.5, 0.5)]	1.2 ms	2.1 ms	-42.36%
❌	Simulation	bench_dict_mask[(0.9, 0.1)]	1.2 ms	2 ms	-39.83%
❌	Simulation	bench_dict_mask[(0.1, 0.1)]	1.2 ms	2 ms	-39.82%
❌	Simulation	bench_dict_mask[(0.01, 0.01)]	1.2 ms	2 ms	-38.36%
❌	Simulation	bench_dict_mask[(0.9, 0.5)]	1.2 ms	2.1 ms	-42.38%
❌	Simulation	bench_dict_mask[(0.1, 0.5)]	1.2 ms	2.1 ms	-42.39%
⚡	Simulation	patched_take_200k_dispersed	5.6 ms	5.1 ms	+11.29%

---

_{Comparing ji/mask-compute-fn-remove (af76682) with develop (d9d75da)}

Footnotes

1. 1268 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[robert3005]

this seems weird

[joseph-isaacs]

this handles it in the outer fn, I guess we could panic here?

[robert3005]

I would panic

[robert3005]

I think you also need to change benchmarks to recursively canonicalize