Proves NP-hardness of distinguishing value 1-δ vs at most 1/q^{1-ε} for 2P1R projection games with large alphabet q (power of 2), improving the alphabet-soundness tradeoff over prior work.
11By(1, q δ2ℓη)-pseudo-random inZoom[A, B]we mean thatL ′ does not increase its fractional size toq δ2ℓηwhen restricted to any zoom-in containingAor any zoom-out contained inB
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CC 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Near Optimal Alphabet-Soundness Tradeoff PCPs
Proves NP-hardness of distinguishing value 1-δ vs at most 1/q^{1-ε} for 2P1R projection games with large alphabet q (power of 2), improving the alphabet-soundness tradeoff over prior work.