Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit...
-
Upload
oscar-hutchinson -
Category
Documents
-
view
222 -
download
2
Transcript of Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit...
![Page 1: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/1.jpg)
Better Pseudorandom Generators from Milder Pseudorandom
Restrictions
Raghu Meka (IAS)Parikshit Gopalan, Omer Reingold
(MSR-SVC) Luca Trevian (Stanford), Salil Vadhan (Harvard)
![Page 2: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/2.jpg)
Can we generate random bits?
![Page 3: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/3.jpg)
Can we generate random bits?
![Page 4: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/4.jpg)
Pseudorandom Generators
Stretch bits to fool a class of “test functions” F
![Page 5: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/5.jpg)
Can we generate random bits?
• Complexity theory, algorithms, streaming
• Strong positive evidence: hardness vs randomness – NW94, IW97, …
• Unconditionally? Duh.
![Page 6: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/6.jpg)
Can we generate random bits?
• Restricted models: bounded depth circuits (AC0), bounded space algorithms
Nis91, Bazzi09, B10, … Nis90, NZ93, INW94, …
![Page 7: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/7.jpg)
•
Reference Seed-length
Nisan 91
LVW 93
Bazzi 09
DETT 10
DETT 10
PRGs for AC0
For polynomially small error best waseven for read-once CNFs.
![Page 8: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/8.jpg)
PRGs for Small-space
Reference Seed-length
Nisan 90, INW 94
Lu 01
BRRY10, BV10, KNP11, De11
For polynomially small error best waseven for comb. rectangles.
![Page 9: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/9.jpg)
This Work
PRGs with polynomial small error
![Page 10: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/10.jpg)
Why Small Error?
• Because we “should” be able to
• Symptomatic: const. error for large depth implies poly. error for smaller depth
• Applications: algorithmic derandomizations, complexity lowerbounds
![Page 11: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/11.jpg)
This Work
Generic new technique: iterative application of mild random
restrictions.
1. PRG for comb. rectangles with seed .
2. PRG for read-once CNFs with seed .
3. HSG for width 3 branching programs with seed .
![Page 12: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/12.jpg)
Combinatorial Rectangles
Applications: Number theory, analysis, integration, hardness amplification
![Page 13: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/13.jpg)
PRGs for Comb. Rectangles
Small set preserving volume
Volume of rectangle ~ Fraction of positive PRG points
![Page 14: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/14.jpg)
Thm: PRG for comb. rectangles with seed .
PRGs for Combinatorial Rectangles
Reference Seed-lengthEGLNV92
LLSZ93
ASWZ96
Lu01
![Page 15: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/15.jpg)
Read-Once CNFs
Each variable appears at most once
Thm: PRG for read-once CNFs with seed .
![Page 16: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/16.jpg)
This Talk
Comb. Rectangles similar but different
Thm: PRG for read-once CNFs with seed .
![Page 17: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/17.jpg)
Outline
1. Main generator: mild (pseudo)random restrictions.
2. Interlude: Small-bias spaces, Tribes
3. Analysis: variance dampening, approximating symmetric functions.
The “real” stuff happens here.
![Page 18: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/18.jpg)
Random Restrictions
• Switching lemma – Ajt83, FSS84, Has86
* * *1 100 0 0** *** *
![Page 19: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/19.jpg)
• Problem: No strong derandomized switching lemmas.
PRGs from Random Restrictions
• AW85: Use “pseudorandom restrictions”.
* * ** *** * *
![Page 20: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/20.jpg)
* * * * * ** * * 0 0 1 0 0 00 0 0
Mild Psedorandom Restrictions
• Restrict half the bits (pseudorandomly).
* * * * * *
“Simplification”: Can be fooled by small-bias spaces.
* * *
![Page 21: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/21.jpg)
Thm: PRG for read-once CNFs with seed .
Repeat Randomness:
Full Generator Construction
Pick half using almost k-wise
* * * * * * * *
Small-bias
* * * *
Small-bias
* *
Small-bias
![Page 22: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/22.jpg)
Outline
1. Main generator: mild (pseudo)-random restrictions.
2. Interlude: Small-bias spaces, Tribes
3. Analysis: variance dampening, approximating symmetric functions.
![Page 23: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/23.jpg)
Toy example: Tribes
Read-once CNF and a Comb. Rectangle
![Page 24: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/24.jpg)
Small-bias Spaces
• Fundamental objects in pseudorandomness
• NN93, AGHP92: can sample with bits
![Page 25: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/25.jpg)
Small-bias Spaces
• PRG with seed • Tight: need bias
![Page 26: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/26.jpg)
The “real” stuff happens here.
Outline
1. Main generator: mild (pseudo)-random restrictions.
2. Interlude: Small-bias spaces, Tribes
3. Analysis: variance dampening, approximating symmetric functions.
![Page 27: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/27.jpg)
Analysis Sketch
Pick half using almost k-wise
* * * * * * * *
Small-bias
* * * *
Small-bias
* *
Small-bias
* * * * * * * *
Uniform
1. Error is small2. Size reduces:
![Page 28: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/28.jpg)
Main idea: Average over uniform to study “bias function”.
• First try: fix uniform bits (averaging argument)
• Problem: still Tribes
0 1 0 0 0 10 0 0
Pick half using almost k-wise
* * * * * ** * *
Analysis for Tribes
* * * * * ** * * * * * * * ** * *Pick exactly half from each clause
White = small-biasYellow = uniform
* * * * * ** * * 0 1 0 0 0 10 0 0
![Page 29: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/29.jpg)
Fooling Bias Functions
• Fix a read-once CNF f. Want:
• Define bias function: False if we fixed X!
![Page 30: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/30.jpg)
Fooling Bias Functions• Let
![Page 31: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/31.jpg)
Fooling Bias Functions
“Variance dampening”: makes things work.
(Without “dampening”)
1−2−𝑤
![Page 32: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/32.jpg)
Fooling Bias Functions
: ’th symmetric polynomial
• F’s fooled by small-bias• ’s decrease geometrically under uniform• No such decrease for small-bias• Conditional decrease: decrease
conditioned on a high probability event (cancellations happen)
![Page 33: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/33.jpg)
Ex: If then
An Inequality for Symmetric Polynomials
Lem:
Proof uses Newton-Girard identities.
Comes from variance dampening.
![Page 34: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/34.jpg)
Summary1. Main generator: mild (pseudo)-
random restrictions.
2. Small-bias spaces and Tribes
3. Analysis: variance dampening, approximating sym. functions.
PRG for RCNFs
Combinatorial rectangles similar but different
![Page 35: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/35.jpg)
Open Problems
Q: Use techniques for other classes? Small-space?
•
![Page 36: Better Pseudorandom Generators from Milder Pseudorandom Restrictions Raghu Meka (IAS) Parikshit Gopalan, Omer Reingold (MSR-SVC) Luca Trevian (Stanford),](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649ce45503460f949b17df/html5/thumbnails/36.jpg)
Thank you
“The best throw of the die is to throw it away”
-