Create MLE eval verifier component

[andrewmilson]

This change is 

[andrewmilson]

• Create blake example that uses GKR for lookups #807
• Create MLE eval verifier component #805 👈
• Create MLE eval prover component #804
• Add build trace functions for MLE eval component #803
• Remove InteractionElements and LookupValues #802
• Pass entire mask to components #801
• Add MleCollection for accumulating MLEs #800
• Combine eq step constraints into single constraint #799
• Move all MLE eval related constraints into same module #798
• dev

This stack of pull requests is managed by Graphite. Learn more about stacking.

Join @andrewmilson and the rest of your teammates on Graphite

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.47%. Comparing base (cd1648e) to head (cc6929c).

Additional details and impacted files

@@           Coverage Diff           @@
##              dev     #805   +/-   ##
=======================================
  Coverage   92.47%   92.47%           
=======================================
  Files          90       90           
  Lines       12400    12400           
  Branches    12400    12400           
=======================================
  Hits        11467    11467           
  Misses        825      825           
  Partials      108      108           

Flag	Coverage Δ
	92.47% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[shaharsamocha7]

Reviewed 1 of 1 files at r2, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @andrewmilson and @spapinistarkware)

---

crates/prover/src/examples/xor/gkr_lookups/mle_eval.rs line 356 at r2 (raw file):

        let mle_coeff_col_eval = self.mle_coeff_column_oracle.evaluate_at_point(point, mask);
        let carry_quotients_col_eval = eval_carry_quotient_col(&self.mle_eval_point, point);
        let is_first = eval_is_first(trace_coset, point);

Should we use this succinct version instead of committing to the is_first column in the constant tree?

Code quote:

eval_is_first(trace_coset, point)

[andrewmilson]

Reviewable status: 0 of 1 files reviewed, 1 unresolved discussion (waiting on @shaharsamocha7 and @spapinistarkware)

---

crates/prover/src/examples/xor/gkr_lookups/mle_eval.rs line 356 at r2 (raw file):

Previously, shaharsamocha7 wrote…

Should we use this succinct version instead of committing to the is_first column in the constant tree?

We could. I remember ShaharP mentioning he thought constant column might be more efficient for the verifier. Not sure if this is just Bitcoin case or all cases

[shaharsamocha7]

Reviewed 1 of 1 files at r3, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @andrewmilson and @spapinistarkware)

---

crates/prover/src/examples/xor/gkr_lookups/mle_eval.rs line 356 at r2 (raw file):

Previously, andrewmilson (Andrew Milson) wrote…

We could. I remember ShaharP mentioning he thought constant column might be more efficient for the verifier. Not sure if this is just Bitcoin case or all cases

coset_vanishing(coset, p) / (point_vanishing(coset.initial, p) * BaseField::from(1 << coset.log_size))

Maybe it is because of the inverse but I guess that we can implement it efficiently using a hint.