From rmhulett at stanford.edu Thu Nov 1 13:19:05 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Thu, 1 Nov 2018 20:19:05 +0000
Subject: [theoryseminar] Theory Lunch  Andrew Stolman (UCSC)
InReplyTo:
References:
MessageID: <2ae6d42651ed4e32882ea778ed447941@email.android.com>
Someone left a glasses case in 463A, please claim!
On Oct 30, 2018 1:28 PM, Reyna Marie Hulett wrote:
Hi everybody,
This week's theory lunch will be given by Andrew Stolman (UCSC) on ?Finding forbidden minors through random walks: an almost optimal onesided tester for minorclosed properties?see abstract below. As always, please join us Thursday from noon to 1 pm in Gates 463A!

Abstract:
Let [G] be an undirected, bounded degree graph with vertices. Fix a finite graph [H] , and suppose one must remove edges from [G] to make it [H] minorfree (for some small constant [\varepsilon > 0] ). We give an n1/2 + o(1)time randomized algorithm that, with high probability, finds an [H] minor in such a graph. As an application, suppose one must remove edges from a bounded degree graph [G] to make it planar. This result implies an algorithm, with the same running time, that produces a [K_{3,3}] or [K_5] minor in [G] . No prior sublinear time bound was known for this problem.
By the graph minor theorem, we get an analogous result for any minorclosed property. Up to [n^{o(1)}] factors, this resolves a conjecture of BenjaminiSchrammShapira (STOC 2008) on the existence of onesided property testers for minorclosed properties. Furthermore, our algorithm is nearly optimal, by an [\Omega(\sqrt{n})] lower bound of Czumaj et al (RSA 2014).
This result appeared in FOCS 2018 and the paper can be found here https://arxiv.org/abs/1805.08187.
Joint work with C. Seshadhri of UCSC and Akash Kumar of Purdue University.

Cheers,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
 next part 
A nontext attachment was scrubbed...
Name: 20181101_131751.jpg
Type: image/jpeg
Size: 1330517 bytes
Desc: 20181101_131751.jpg
URL:
From rmhulett at stanford.edu Thu Nov 1 13:19:05 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Thu, 1 Nov 2018 20:19:05 +0000
Subject: [theoryseminar] Theory Lunch  Andrew Stolman (UCSC)
InReplyTo:
References:
MessageID: <2ae6d42651ed4e32882ea778ed447941@email.android.com>
Someone left a glasses case in 463A, please claim!
On Oct 30, 2018 1:28 PM, Reyna Marie Hulett wrote:
Hi everybody,
This week's theory lunch will be given by Andrew Stolman (UCSC) on ?Finding forbidden minors through random walks: an almost optimal onesided tester for minorclosed properties?see abstract below. As always, please join us Thursday from noon to 1 pm in Gates 463A!

Abstract:
Let [G] be an undirected, bounded degree graph with vertices. Fix a finite graph [H] , and suppose one must remove edges from [G] to make it [H] minorfree (for some small constant [\varepsilon > 0] ). We give an n1/2 + o(1)time randomized algorithm that, with high probability, finds an [H] minor in such a graph. As an application, suppose one must remove edges from a bounded degree graph [G] to make it planar. This result implies an algorithm, with the same running time, that produces a [K_{3,3}] or [K_5] minor in [G] . No prior sublinear time bound was known for this problem.
By the graph minor theorem, we get an analogous result for any minorclosed property. Up to [n^{o(1)}] factors, this resolves a conjecture of BenjaminiSchrammShapira (STOC 2008) on the existence of onesided property testers for minorclosed properties. Furthermore, our algorithm is nearly optimal, by an [\Omega(\sqrt{n})] lower bound of Czumaj et al (RSA 2014).
This result appeared in FOCS 2018 and the paper can be found here https://arxiv.org/abs/1805.08187.
Joint work with C. Seshadhri of UCSC and Akash Kumar of Purdue University.

Cheers,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
 next part 
A nontext attachment was scrubbed...
Name: 20181101_131751.jpg
Type: image/jpeg
Size: 1330517 bytes
Desc: 20181101_131751.jpg
URL:
From reingold at stanford.edu Thu Nov 1 14:58:46 2018
From: reingold at stanford.edu (Omer Reingold)
Date: Thu, 1 Nov 2018 14:58:46 0700
Subject: [theoryseminar] TOCASV in a week (Friday Nov. 9)!
MessageID:
Dear Friends and Colleagues,
Appended below is the schedule and abstracts for the TOCASV meeting next
week that will be hosting Shafi Goldwasser for the Motwani colloquium.
The event will take place on Friday November 9 at Huang Mackenzie center
room 300. For direction and parking instructions please see
https://tfsa.stanford.edu/pdfs/Parking_and_Directions.pdf.
We are looking forward to seeing you all there!
The organizing committee: Moses Charikar (Stanford), Ravi Kumar (Google),
Omer Reingold (Stanford), D. Sivakumar (Google) and Gregory Valiant
(Stanford)
11:0011:45 Avishay Tal, Simons Institute & Stanford University,
Oracle Separation of BQP and the Polynomial
Hierarchy
11:4512:30 Badih Ghazi, Google
ResourceEfficient Common Randomness and
Secret Key Generation
12:302:00 lunch
2:003:15 Motwani colloquium, Shafi Goldwasser, Simons Institute,
UC Berkeley
Pseudo Deterministic Algorithms and Proofs
3:155:30 Happy hour and student talks
Abstracts:
Avishay Tal, *Oracle Separation of BQP and the Polynomial Hierarchy*
In their seminal paper, Bennett, Bernstein, Brassard and Vazirani [SICOMP,
1997] showed that relative to an oracle, quantum algorithms are unable to
solve NPcomplete problems in subexponential time (i.e., that Grover's
search is optimal in this setting).
In this work, we show a strong converse to their result. Namely, we show
that, relative to an oracle, there exist computational tasks that can be
solved efficiently by a quantum algorithm, but require exponential time for
any algorithm in the polynomial hierarchy (a hierarchy of complexity
classes that captures P, NP, coNP, and their generalizations).
The tasks that exhibit this "quantum advantage" arise from a
pseudorandomness approach initiated by Aaronson [STOC, 2010]. Our core
technical result is constructing a distribution over Boolean strings that
"look random" to constantdepth circuits of quasipolynomial size, but can
be distinguished from the uniform distribution by very efficient quantum
algorithms.
Joint work with Ran Raz
Badih Ghazi, *ResourceEfficient Common Randomness and Secret Key
Generation*
The task of manipulating randomness has been a subject of intense
investigation in computational complexity with dispersers, extractors,
pseudorandom generators, condensers, mergers being just a few of the
objects of interest. All these tasks consider a single processor massaging
random samples from an unknown source.
In this talk, I will discuss a less studied setting where randomness is
distributed among different players who would like to convert it to other
forms in an efficient manner and with little communication. For instance
players may be given access to a source of biased correlated bits and their
goal may be to get a common random string out of this source. Even the
setting where the source is known leads to some interesting questions that
have been explored since the 70s with striking constructions and some
surprisingly hard questions. After giving some background, I will describe
recent work which explores the task of extracting common randomness from
correlated sources with bounds on either the sample complexity or on the
number of rounds of interaction.
Based on joint works with T.S. Jayram, Mitali Bafna, Noah Golowich and
Madhu Sudan
Shafi Goldwasser, *Pseudo Deterministic Algorithms and Proofs*
Probabilistic algorithms for both decision and search problems can offer
significant complexity improvements over deterministic algorithms. One
major difference, however, is that they may output different solutions for
different choices of randomness. This makes correctness amplification
impossible for search algorithms and is less than desirable in setting
where uniqueness of output is important such as generation of system wide
cryptographic parameters or distributed setting where different sources of
randomness are used. Pseudodeterministic algorithms are a class of
randomized search algorithms, which output a unique answer with high
probability. Intuitively, they are indistinguishable from deterministic
algorithms by an polynomial time observer of their input/output behavior.
In this talk I will describe what is known about pseudodeterministic
algorithms in the sequential, sublinear and parallel setting. We will also
describe n extension of pseudodeterministic algorithms to interactive
proofs for search problems where the veri
fier is guaranteed with high probability to output the same output on
different executions, regardless of the prover strategies.
 next part 
An HTML attachment was scrubbed...
URL:
From gvaliant at cs.stanford.edu Fri Nov 2 07:45:49 2018
From: gvaliant at cs.stanford.edu (Gregory Valiant)
Date: Fri, 2 Nov 2018 07:45:49 0700
Subject: [theoryseminar] postdeadline happy hour today @6pm!
MessageID:
Hi Friends,
We'll have our usual postdeadline happy hour, today at 6pm, in gates
463a. I probably won't have time to go to the store beforehand, so the
material sustenance on offer might be a bit sparse, but we will have plenty
of food for the mind and/or soul. (Though if we all bring whatever beer is
hiding in our offices, I imagine we should be all set : )
Hope to see you this evening,
g
 next part 
An HTML attachment was scrubbed...
URL:
From ccanonne at cs.stanford.edu Fri Nov 2 07:50:47 2018
From: ccanonne at cs.stanford.edu (=?UTF8?Q?Cl=c3=a9ment_Canonne?=)
Date: Fri, 2 Nov 2018 07:50:47 0700
Subject: [theoryseminar] postdeadline happy hour today @6pm!
InReplyTo:
References:
MessageID:
On 11/2/18 7:45 AM, Gregory Valiant wrote:
> (Though if we all bring whatever beer is hiding in our offices, I
> imagine we should be all set : )
I confirm :)
Also, important note: having submitted, or attended to submit, a paper
is *not* a prerequisite to attend the happy hour.
Best,
 Cl?ment
From ofirgeri at stanford.edu Fri Nov 2 13:12:46 2018
From: ofirgeri at stanford.edu (Ofir Geri)
Date: Fri, 2 Nov 2018 20:12:46 +0000
Subject: [theoryseminar] Theory Seminar (11/2): Rotem Oshman
InReplyTo:
References:
MessageID:
Reminder: Rotem Oshman's talk is today at 3pm. Please note the nonstandard location (Gates 392).
________________________________
From: Ofir Geri
Sent: Wednesday, October 31, 2018 4:19:35 PM
To: thseminar at cs.stanford.edu
Subject: Theory Seminar (11/2): Rotem Oshman
Hi all,
This Friday in theory seminar, Rotem Oshman (TelAviv University) will give a talk on Two Open Problems in Distributed Graph Algorithms (see abstract below). The talk will be at 3pm in Gates 392  please note the unusual location!
Hope to see you there!
Ofir
Two Open Problems in Distributed Graph Algorithms
Speaker: Rotem Oshman (TelAviv University)
In a distributed graph algorithm, we have a network of computing nodes, where each node initially knows only its own local neighborhood; the nodes communicate over the network edges in order to solve some problem on the network graph. We are interested in algorithms that are fast, but also do not require a lot of communication between the network nodes.
In this talk I will describe recent algorithms and lower bounds for two graph problems: exact maximum bipartite matching, and testing whether the network contains an evenlength cycle of a specific length. Both problems do not have matching upper and lower bounds, so their complexity remains open. The talk will not assume any prior knowledge about distributed computing.
 next part 
An HTML attachment was scrubbed...
URL:
From moses at cs.stanford.edu Mon Nov 5 13:15:03 2018
From: moses at cs.stanford.edu (Moses Charikar)
Date: Mon, 5 Nov 2018 13:15:03 0800
Subject: [theoryseminar] Nima Anari at probability seminar today, 4pm
MessageID:
Hi folks,
Nima is speaking at the probability seminar today at 4pm on logconcave
polynomials. Details below.
Best,
Moses
Monday, November 5, 4pm: Probability Seminar, Sequoia Hall Room 200
Nima Anari, Stanford Computer Science
Title: LogConcave Polynomials: Counting and Sampling Matroid Bases and
Beyond
Abstract:
I will discuss an analytic property of multivariate polynomials, which we
call complete logconcavity. This property defines a large class of
discrete distributions that should be thought of as the discrete analog of
the wellstudied continuous logconcave distributions. Examples satisfying
this property include determinantal point processes and fractional powers
of them, the random cluster model and Potts model for some regime of
parameters, strongly Rayleigh distributions, uniform distributions over
bases or independent sets of matroids, and other generalizations. Complete
logconcavity is intimately connected to HodgeRiemann relations of order
one from Hodge theory, and also surprisingly high dimensional expanders.
The second connection allows us to efficiently sample from discrete
distributions satisfying complete logconcavity. For many of the example
distributions, this provides the first efficient sampling algorithm. In
particular we obtain the first polynomial time algorithm to sample
approximately uniformly from bases of an arbitrary matroid.
This is joint work with Kuikui Liu, Shayan Oveis Gharan, and Cynthia
Vinzant.
 next part 
An HTML attachment was scrubbed...
URL:
From rmhulett at stanford.edu Mon Nov 5 14:08:23 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Mon, 5 Nov 2018 22:08:23 +0000
Subject: [theoryseminar] Theory Lunch  Mohammad Akbarpour
MessageID:
Hi all,
This week at theory lunch, Mohammad Akbarpour will present "Just a Few Seeds More: Value of Network Information for Diffusion"see abstract below. As always, please join us Thursday from noon to 1 pm in Gates 463A!

Abstract:
Identifying the optimal set of individuals to first receive information (?seeds?) in a social network is a widelystudied question in many settings, such as the diffusion of information, microfinance programs, and new technologies. Numerous studies have proposed various networkcentrality based heuristics to choose seeds in a way that is likely to boost diffusion. Here we show that, for some frequently studied diffusion processes, randomly seeding s+x individuals can prompt a larger cascade than optimally targeting the best s individuals, for a small x. We prove our results for large classes of random networks, but also show that they hold in simulations over several realworld networks. This suggests that the returns to collecting and analyzing network information to identify the optimal seeds may not be economically significant. Given these findings, practitioners interested in communicating a message to a large number of people may wish to compare the cost of networkbased targeting to that of slightly expanding initial outreach.
Available at https://web.stanford.edu/~mohamwad/NetworkSeeding.pdf

Cheers,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
From rmhulett at stanford.edu Mon Nov 5 14:08:23 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Mon, 5 Nov 2018 22:08:23 +0000
Subject: [theoryseminar] Theory Lunch  Mohammad Akbarpour
MessageID:
Hi all,
This week at theory lunch, Mohammad Akbarpour will present "Just a Few Seeds More: Value of Network Information for Diffusion"see abstract below. As always, please join us Thursday from noon to 1 pm in Gates 463A!

Abstract:
Identifying the optimal set of individuals to first receive information (?seeds?) in a social network is a widelystudied question in many settings, such as the diffusion of information, microfinance programs, and new technologies. Numerous studies have proposed various networkcentrality based heuristics to choose seeds in a way that is likely to boost diffusion. Here we show that, for some frequently studied diffusion processes, randomly seeding s+x individuals can prompt a larger cascade than optimally targeting the best s individuals, for a small x. We prove our results for large classes of random networks, but also show that they hold in simulations over several realworld networks. This suggests that the returns to collecting and analyzing network information to identify the optimal seeds may not be economically significant. Given these findings, practitioners interested in communicating a message to a large number of people may wish to compare the cost of networkbased targeting to that of slightly expanding initial outreach.
Available at https://web.stanford.edu/~mohamwad/NetworkSeeding.pdf

Cheers,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
From moses at cs.stanford.edu Tue Nov 6 22:26:02 2018
From: moses at cs.stanford.edu (Moses Charikar)
Date: Tue, 6 Nov 2018 22:26:02 0800
Subject: [theoryseminar] Motwani CS Theory Colloquium: Shafi Goldwasser
(Nov 9)
MessageID:
The Motwani Distinguished Lectures are a series of theory colloquia aimed
at a broad audience. The next lecture in the series will be given this
Friday, Nov 9, by Shafi Goldwasser, Turing award winner and director of the
Simons Institute at Berkeley, on Pseudo Deterministic Algorithms and Proofs
(abstract below). You should definitely attend if you can!
The talk is at 2pm on Nov 9 in the Mackenzie Room (Huang Engineering
Center, Room 300).
For direction and parking instructions please see
https://tfsa.stanford.edu/pdfs/Parking_and_Directions.pdf.
There will be a happy hour immediately following the talk. This Motwani
Colloquium is part of TOCASV, our biannual gathering of CS theoreticians
in academia and industry in the Silicon Valley area. See the full schedule
here:
https://sites.google.com/view/tocasvdaynov2018/
Hope to see you there!
Cheers,
Moses
Title: Pseudo Deterministic Algorithms and Proofs
Shafi Goldwasser
Abstract
Probabilistic algorithms for both decision and search problems can offer
significant complexity improvements over deterministic algorithms. One
major difference, however, is that they may output different solutions for
different choices of randomness. This makes correctness amplification
impossible for search algorithms and is less than desirable in setting
where uniqueness of output is important such as generation of system wide
cryptographic parameters or distributed setting where different sources of
randomness are used. Pseudodeterministic algorithms are a class of
randomized search algorithms, which output a unique answer with high
probability. Intuitively, they are indistinguishable from deterministic
algorithms by a polynomial time observer of their input/output behavior. In
this talk I will describe what is known about pseudodeterministic
algorithms in the sequential, sublinear and parallel setting. We will also
describe an extension of pseudodeterministic algorithms to interactive
proofs for search problems where the verifier is guaranteed with high
probability to output the same output on different executions, regardless
of the prover strategies.
 next part 
An HTML attachment was scrubbed...
URL:
From ccanonne at cs.stanford.edu Wed Nov 7 17:01:56 2018
From: ccanonne at cs.stanford.edu (=?UTF8?Q?Cl=c3=a9ment_Canonne?=)
Date: Wed, 7 Nov 2018 17:01:56 0800
Subject: [theoryseminar] TCS+ talk: Wednesday, November 14, Urmila Mahadev,
UC Berkeley
InReplyTo: <7c7f1a3e747b9fc483179c5514ccf03a@cs.columbia.edu>
References: <7c7f1a3e747b9fc483179c5514ccf03a@cs.columbia.edu>
MessageID: <27644919b306dcb3a1886b8ce203bf9c@cs.stanford.edu>
Hi everyone,
Next Wednesday morning, we'll have a TCS+ seminar at 10am, where Urmila
Mahadev from UC Berkeley will (remotely) tell us about her recent work
on "Classical Homomorphic Encryption for Quantum Circuits" (abstract
below).
As usual, this will be at *10am* in Gates 463; as usual we'll be able
to ask questions during the talk, and as usual there will be pastries
and light breakfast, from 9:55 until disappearance.
See you then,
 Cl?ment

Speaker: Urmila Mahadev (UC Berkeley)
Title: Classical Homomorphic Encryption for Quantum Circuits
Abstract: We present the first leveled fully homomorphic encryption
scheme for quantum circuits with classical keys. The scheme allows a
classical client to blindly delegate a quantum computation to a quantum
server: an honest server is able to run the computation while a
malicious server is unable to learn any information about the
computation. We show that it is possible to construct such a scheme
directly from a quantum secure classical homomorphic encryption scheme
with certain properties. Finally, we show that a classical homomorphic
encryption scheme with the required properties can be constructed from
the learning with errors problem.
From tengyuma at stanford.edu Mon Nov 12 09:48:15 2018
From: tengyuma at stanford.edu (Tengyu Ma)
Date: Mon, 12 Nov 2018 09:48:15 0800
Subject: [theoryseminar] potentially interesting talk at stats seminar
MessageID:
Hi theory folks,
FYI, from the abstract, it sounds related to graph algorithms and
algorithmic game theory; so I thought maybe somebody would be interested
https://statistics.stanford.edu/sites/default/files/Nov132018.pdf
Best,
Tengyu

Sent from my phone. Apologies for brevity.
 next part 
An HTML attachment was scrubbed...
URL:
From marykw at stanford.edu Mon Nov 12 14:50:16 2018
From: marykw at stanford.edu (Mary Wootters)
Date: Mon, 12 Nov 2018 14:50:16 0800
Subject: [theoryseminar] Fwd: Summer internship at Google research NYC
InReplyTo:
References:
MessageID:
Hi all,
See below for an advertisement about summer internships at Google Research
NY.
Mary
 Forwarded message 
From: Srinadh Bhojanapalli
Date: Mon, Nov 12, 2018 at 1:40 PM
Subject: Summer internship at Google research NYC
To:
Hi Mary,
I hope you are doing well. I have recently joined Google research in NYC.
We have a summer internship program focused on algorithms and machine
learning at Google research NY. The official call and the application
procedure is given below. It will be great if you can circulate it among
your students and other faculty.
Thanks,
srinadh.
Google Research NY: Internships for Summer 2019
A limited number of summer internship positions are available at Google
Research, New York in the areas of machine learning, optimization, deep
learning, computer vision and related fields. We seek motivated and
enthusiastic colleagues with a solid research agenda and a demonstrated
ability to generate new ideas and innovate.
The candidates will have an opportunity to work closely with many
researchers at Google, and have access to Google's unique computing
infrastructure. The candidates will be expected to produce highquality
research and will be encouraged to publish in top conferences and journals
in their areas of expertise.
Applicants must

have strong theoretical and analytical foundations;

be enrolled in a Ph.D. program;

have good prototyping skills.
The internship is planned for the summer of 2019. A minimum of 12 weeks is
required.
Applicants should send a curriculum vitae (including a list of
publications) and contact information of two references who can comment on
the applicant's research potential, one of whom should be the dissertation
advisor, to *applyresearchinternshipny2019 at google.com
*. The application deadline
is *December 14, 2018* but the selection process will start on firstcome,
firstserved basis.
 next part 
An HTML attachment was scrubbed...
URL:
From tengyuma at stanford.edu Mon Nov 12 21:22:28 2018
From: tengyuma at stanford.edu (Tengyu Ma)
Date: Mon, 12 Nov 2018 21:22:28 0800
Subject: [theoryseminar] [csseminars] Theory Seminar: Andrew ChiChih Yao
on Auction and Blockchain
MessageID:
Hi everyone,
We will have a special theory seminar by the Turing Award Laureate Prof.
Andrew ChiChih Yao from Tsinghua University at 2PM3PM this Friday, Nov
16th, in Gates 104. Hope to see you there!
*Title:* Some Perspectives on FinTech
*Speaker:* Andrew ChiChih Yao, Tsinghua University
*Abstract:* FinTech can be seen as the meeting of minds between economics
and computer science in the digital age. One of its major intellectual
foundations is the interdisciplinary subject of multiparty computation,
which involves reliable distributed computing and cryptography from the
side of computer science, and efficient mechanism design for financial
activities from the side of economics. In this talk we discuss some recent
work in auction and blockchain from this perspective. For example, is it
true that more revenue can always be extracted from an auction where the
bidders are more willing to pay than otherwise? Can more revenue be
extracted when the bidders are more risktolerant than otherwise? We also
present some new results on blockchain fees. These results help shed light
on some structural questions in economics whose answers are nonobvious.
*Bio* (from Wiki Award section): Andrew ChiChih Yao is currently a
Professor and the Dean of Institute for Interdisciplinary Information
Sciences (IIIS) at Tsinghua University. In 1996 he was awarded the Knuth
Prize. He received the Turing Award, the most prestigious award in computer
science, in 2000, "in recognition of his fundamental contributions to the
theory of computation, including the complexitybased theory of
pseudorandom number generation, cryptography, and communication
complexity". He is a member of U.S. National Academy of Sciences, a fellow
of the American Academy of Arts and Sciences, a fellow of the American
Association for the Advancement of Science, a fellow of the Association for
Computing Machinery, and an academician of Chinese Academy of Sciences.

Sent from my phone. Apologies for brevity.
 next part 
An HTML attachment was scrubbed...
URL:
From ofirgeri at stanford.edu Tue Nov 13 12:25:09 2018
From: ofirgeri at stanford.edu (Ofir Geri)
Date: Tue, 13 Nov 2018 20:25:09 +0000
Subject: [theoryseminar] Theory Seminar (11/16): Jonathan Mosheiff
MessageID:
Hi all,
This Friday, Jonathan Mosheiff from Weizmann will give a theory seminar talk: On the weight distribution of random binary linear codes (abstract below).
The talk will take place in Gates 392 at 3:15 PM (shortly after Andrew ChiChih Yao's talk). Please note the different locations for the two talks.
Jonathan will be visiting both Thursday 11/15 and Friday 11/16, and will also be giving a (different) talk in the combinatorics seminar on 11/15. If you'd like to meet with him, please contact Mary at marykw at stanford.edu
Hope to see you there!
Ofir
On the weight distribution of random binary linear codes
Speaker: Jonathan Mosheiff (Weizmann)
A random (binary) linear code is a dimension lambda*n (0 < \lambda < 1) linear subspace of the binary ndimensional hypercube, chosen uniformly from among all such subspaces. Such codes play an important role in the theory of error correcting codes, since they achieve the best known rate vs. distance tradeoff, i.e., the GilbertVarshamov lower bound. Under a random errors regime, the problem of decoding these codes is known as Learning Parity with Noise, and has many cryptographic applications. This work is motivated by the contrast between the importance of random linear codes and how little we know about them.
Much of the interesting information about a code C is captured by its weight distribution. This is the vector (w_0,w_1,...,w_n) where w_i counts the elements of C with Hamming weight i. In this work we study the weight distribution of random linear codes. In our main result we compute the moments of the random variable w_(gamma*n), where 0 < \gamma < 1 is a fixed constant and n goes to infinity.
This is joint work with Nati Linial.
 next part 
An HTML attachment was scrubbed...
URL:
From ccanonne at cs.stanford.edu Tue Nov 13 15:42:04 2018
From: ccanonne at cs.stanford.edu (=?UTF8?Q?Cl=c3=a9ment_Canonne?=)
Date: Tue, 13 Nov 2018 15:42:04 0800
Subject: [theoryseminar] TCS+ talk: Wednesday, November 14,
Urmila Mahadev, UC Berkeley
InReplyTo: <27644919b306dcb3a1886b8ce203bf9c@cs.stanford.edu>
References: <7c7f1a3e747b9fc483179c5514ccf03a@cs.columbia.edu>
<27644919b306dcb3a1886b8ce203bf9c@cs.stanford.edu>
MessageID: <0e1f8bd522ba8c077e7fafd781d8d0a5@cs.stanford.edu>
Reminder: today is the 13th, so this is happening tomorrow morning!
 Cl?ment
On 11/7/18 5:01 PM, Cl?ment Canonne wrote:
> Hi everyone,
>
> Next Wednesday morning, we'll have a TCS+ seminar at 10am, where Urmila
> Mahadev from UC Berkeley will (remotely) tell us about her recent work
> on? "Classical Homomorphic Encryption for Quantum Circuits" (abstract
> below).
>
> As usual, this will be at *10am* in Gates 463;? as usual we'll be able
> to ask questions during the talk, and as usual there will be pastries
> and light breakfast, from 9:55 until disappearance.
>
> See you then,
>
>  Cl?ment
>
> 
> Speaker: Urmila Mahadev (UC Berkeley)
> Title: Classical Homomorphic Encryption for Quantum Circuits
>
> Abstract: We present the first leveled fully homomorphic encryption
> scheme for quantum circuits with classical keys. The scheme allows a
> classical client to blindly delegate a quantum computation to a quantum
> server: an honest server is able to run the computation while a
> malicious server is unable to learn any information about the
> computation. We show that it is possible to construct such a scheme
> directly from a quantum secure classical homomorphic encryption scheme
> with certain properties. Finally, we show that a classical homomorphic
> encryption scheme with the required properties can be constructed from
> the learning with errors problem.
>
> _______________________________________________
> theoryseminar mailing list
> theoryseminar at lists.stanford.edu
> https://mailman.stanford.edu/mailman/listinfo/theoryseminar
From rmhulett at stanford.edu Wed Nov 14 14:07:59 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Wed, 14 Nov 2018 22:07:59 +0000
Subject: [theoryseminar] Theory Lunch  Yours Truly
MessageID:
Hi folks,
Tomorrow I'll be presenting some work with Ashish Goel and Anilesh Krishnaswamy entitled "Relating Metric Distortion and Fairness of Social Choice Rules." As always, please join us Thursday from noon to 1 pm in Gates 463A.
Note that there is no lunch next weekhave a great Thanksgiving!

Abstract:
The study of voting theory has a long history. However, almost all of that history focused on qualitative, axiomatic ways to evaluate voting rules (e.g., the majority winner criterion, the Condorcet criterion). In contrast, a utilitarian distortion framework offers a quantitative way to evaluate such rules, even though voters submit only a preference ranking over candidates. In this model, we assume the voters have some unknown utility or cost for each candidate, which we can only observe via the preference rankings generated thereby. The distortion of a voting rule is the worstcase approximation ratio of the total social cost of the candidate chosen by that rule versus the optimal candidate (w.r.t. the cost profile). A fairness ratio can analogously be defined as the worstcase approximation ratio for an appropriate, strong measure of fairness (based on approximate majorization).
In this work, we focus on the model where voters' costs for candidates correspond to distances in some arbitrary, unknown metric space. Surprisingly, this modest assumption allows us to obtain constant factor distortion and fairness ratio, e.g., using the Copeland rule. (By contrast, plurality only achieves distortion linear in the number of candidates.) We prove that the distinct notions of distortion and fairness ratio are in fact closely linkedwithin an additive factor of 2 for any voting rule. This implies new results for rules where only the distortion was previously known, and arguably strengthens the case for distortion as a way to evaluate voting rules.

Best,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
From rmhulett at stanford.edu Wed Nov 14 14:07:59 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Wed, 14 Nov 2018 22:07:59 +0000
Subject: [theoryseminar] Theory Lunch  Yours Truly
MessageID:
Hi folks,
Tomorrow I'll be presenting some work with Ashish Goel and Anilesh Krishnaswamy entitled "Relating Metric Distortion and Fairness of Social Choice Rules." As always, please join us Thursday from noon to 1 pm in Gates 463A.
Note that there is no lunch next weekhave a great Thanksgiving!

Abstract:
The study of voting theory has a long history. However, almost all of that history focused on qualitative, axiomatic ways to evaluate voting rules (e.g., the majority winner criterion, the Condorcet criterion). In contrast, a utilitarian distortion framework offers a quantitative way to evaluate such rules, even though voters submit only a preference ranking over candidates. In this model, we assume the voters have some unknown utility or cost for each candidate, which we can only observe via the preference rankings generated thereby. The distortion of a voting rule is the worstcase approximation ratio of the total social cost of the candidate chosen by that rule versus the optimal candidate (w.r.t. the cost profile). A fairness ratio can analogously be defined as the worstcase approximation ratio for an appropriate, strong measure of fairness (based on approximate majorization).
In this work, we focus on the model where voters' costs for candidates correspond to distances in some arbitrary, unknown metric space. Surprisingly, this modest assumption allows us to obtain constant factor distortion and fairness ratio, e.g., using the Copeland rule. (By contrast, plurality only achieves distortion linear in the number of candidates.) We prove that the distinct notions of distortion and fairness ratio are in fact closely linkedwithin an additive factor of 2 for any voting rule. This implies new results for rules where only the distortion was previously known, and arguably strengthens the case for distortion as a way to evaluate voting rules.

Best,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
From ccanonne at cs.stanford.edu Wed Nov 14 14:52:56 2018
From: ccanonne at cs.stanford.edu (=?UTF8?Q?Cl=c3=a9ment_Canonne?=)
Date: Wed, 14 Nov 2018 14:52:56 0800
Subject: [theoryseminar] Upcoming workshop at Simons: Sublinear Algorithms
and NearestNeighbor Search
MessageID: <45bfc130cef2c111d0bfcd62554a7d47@cs.stanford.edu>
Hi everyone,
The week leading to the Theory Retreat, the Simons institute will be
hosting a workshop on "Sublinear Algorithms and NearestNeighbor
Search", which may be of interest to people here:
https://simons.berkeley.edu/datascience20183
It's free, though registration is recommended. The only drawback is that
getting to Berkeley for 9am involves a certain amount of traffic.
Best,

 Cl?ment
From marykw at stanford.edu Thu Nov 15 17:54:29 2018
From: marykw at stanford.edu (Mary Wootters)
Date: Thu, 15 Nov 2018 17:54:29 0800
Subject: [theoryseminar] Course Announcement: EE387/CS250 (Algebraic Error
Correcting Codes)
MessageID:
Hi all,
In the winter I will be teaching EE387/CS250, Algebraic Error Correcting
Codes. The course meets Monday and Wednesday, 4:305:50pm. The official
blurb is below, but the informal pitch for the course is as follows:
This class is an introduction to algebraic error correcting codes (ECCs),
with an emphasis on the diverse applications of ECCs. To define the terms
in the title, an "error correcting code" is a tool to encode data to
protect it from noise, and the adjective "algebraic" means that we will
construct these codes mostly using tools from abstract algebra; the main
workhorse is (variants of) polynomial interpolation over finite fields.
(But don't worry if you're not an expert in abstract algebra! The class is
selfcontained and we will cover all the algebra that we need.)
It turns out that these algebraic tools are extremely useful, and not just
for the "standard" ECC applications of communication and storage. We'll
also talk about applications in cryptography, algorithm design, group
testing, compressed sensing, complexity theory and pseudorandomness. The
goals of the course are to (a) introduce you to the basics of error
correcting codes, and (b) to familiarize you with an algebraic toolkit that
you can use in a variety of settings, not just for ECCs.
The class will involve three problem sets and a final project. The project
can be pretty much anything you want: in the past, some people have done
original research, some have read and digested a few research papers, some
have looked into reallife instances of error correcting codes, and some
have implemented something. (I do have an "art project" option on the
syllabus, but so far no one has taken me up on it...)
If you are interested, please enroll!
Best,
Mary

Official blurb:
*Course Description: *Introduction to the theory of error correcting codes,
emphasizing algebraic constructions and diverse applications throughout
computer science and engineering. Topics include basic bounds on error
correcting codes; ReedSolomon and ReedMuller codes; listdecoding,
listrecovery and locality. Applications may include communication,
storage, complexity theory, pseudorandomness, cryptography, streaming
algorithms, group testing, and compressed sensing.
*Prerequisites:* Linear algebra, basic probability (at the level of, say,
CS109, CME106 or EE178) and "mathematical maturity" (students will be asked
to write proofs). Familiarity with finite fields will be helpful but not
required.
 next part 
An HTML attachment was scrubbed...
URL:
From ofirgeri at stanford.edu Fri Nov 16 12:08:49 2018
From: ofirgeri at stanford.edu (Ofir Geri)
Date: Fri, 16 Nov 2018 20:08:49 +0000
Subject: [theoryseminar] Theory Seminar (11/16): Jonathan Mosheiff
InReplyTo:
References:
MessageID:
Reminder: Two theory seminars today  Andrew ChiChih Yao at 2pm in Gates 104 and Jonathan Mosheiff at 3:15pm in Gates 392.
Please also note that there is no theory lunch or theory seminar next week (Thanksgiving week).
________________________________
From: Ofir Geri
Sent: Tuesday, November 13, 2018 12:25:09 PM
To: thseminar at cs.stanford.edu
Subject: Theory Seminar (11/16): Jonathan Mosheiff
Hi all,
This Friday, Jonathan Mosheiff from Weizmann will give a theory seminar talk: On the weight distribution of random binary linear codes (abstract below).
The talk will take place in Gates 392 at 3:15 PM (shortly after Andrew ChiChih Yao's talk). Please note the different locations for the two talks.
Jonathan will be visiting both Thursday 11/15 and Friday 11/16, and will also be giving a (different) talk in the combinatorics seminar on 11/15. If you'd like to meet with him, please contact Mary at marykw at stanford.edu
Hope to see you there!
Ofir
On the weight distribution of random binary linear codes
Speaker: Jonathan Mosheiff (Weizmann)
A random (binary) linear code is a dimension lambda*n (0 < \lambda < 1) linear subspace of the binary ndimensional hypercube, chosen uniformly from among all such subspaces. Such codes play an important role in the theory of error correcting codes, since they achieve the best known rate vs. distance tradeoff, i.e., the GilbertVarshamov lower bound. Under a random errors regime, the problem of decoding these codes is known as Learning Parity with Noise, and has many cryptographic applications. This work is motivated by the contrast between the importance of random linear codes and how little we know about them.
Much of the interesting information about a code C is captured by its weight distribution. This is the vector (w_0,w_1,...,w_n) where w_i counts the elements of C with Hamming weight i. In this work we study the weight distribution of random linear codes. In our main result we compute the moments of the random variable w_(gamma*n), where 0 < \gamma < 1 is a fixed constant and n goes to infinity.
This is joint work with Nati Linial.
 next part 
An HTML attachment was scrubbed...
URL:
From reingold at stanford.edu Mon Nov 19 09:56:44 2018
From: reingold at stanford.edu (Omer Reingold)
Date: Mon, 19 Nov 2018 09:56:44 0800
Subject: [theoryseminar] Fwd: PCPfest, 1820
InReplyTo:
References:
MessageID:
A great lineup of speakers if you are in the area.
 Forwarded message 
From: Dor Minzer
Date: Mon, Nov 19, 2018, 8:03 AM
Subject: PCPfest, 1820
To:
Dear Omer,
We're holding a PCP centered workshop in TelAviv.
Could you please forward the announcement below
to the relevant lists in Stanford?
Thanks,
Dor
___________________________________
We hereby invite you to the upcoming PCPfest,
to be held on December 1820, 2018 in TelAviv University, Israel.
The workshop will focus on the state of the art theoretical results on
PCPs,
as well as recent applications, in particular, with regards to Blockchain
technology.
More information about the workshop can be found in the website
.
Registration is mandatory (but free).
We would be able to provide limited travel support for participants;
to apply, please indicate it in the registration form.
Looking forward to see you,
The organizers.
 next part 
An HTML attachment was scrubbed...
URL:
From ofirgeri at stanford.edu Sun Nov 25 23:24:46 2018
From: ofirgeri at stanford.edu (Ofir Geri)
Date: Mon, 26 Nov 2018 07:24:46 +0000
Subject: [theoryseminar] Theory Seminar (11/28): Urmila Mahadev
MessageID:
Hi all,
This Wednesday, Urmila Mahadev (UC Berkeley) will be giving a theory seminar talk on Classical Verification of Quantum Computations (abstract below). The talk will be at 3:00 PM in Gates 463A.
Please note that this week's talk is on Wednesday (and there will be no theory seminar on Friday).
Hope to see you there!
Ofir
Classical Verification of Quantum Computations
Speaker: Urmila Mahadev (UC Berkeley)
We present the first protocol allowing a classical computer to interactively verify the result of an efficient quantum computation. We achieve this by constructing a measurement protocol, which allows a classical string to serve as a commitment to a quantum state. The protocol forces the prover to behave as follows: the prover must construct an n qubit state of his choice, nonadaptively measure each qubit in the Hadamard or standard basis as directed by the verifier, and report the measurement results to the verifier. The soundness of this protocol is enforced based on the assumption that the learning with errors problem is computationally intractable for efficient quantum machines.
 next part 
An HTML attachment was scrubbed...
URL:
From rmhulett at stanford.edu Tue Nov 27 12:59:35 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Tue, 27 Nov 2018 20:59:35 +0000
Subject: [theoryseminar] Theory Lunch  Dima Kogan
MessageID:
Hi folks,
In our penultimate theory lunch of the quarter, Dima Kogan will present "From Cryptanalysis to Communication Complexity"see abstract below. As always, please join us Thursday from noon to 1 pm in Gates 463A!

Abstract:
The first problem we will discuss comes from cryptanalysis. In the permutationinversion problem, first studied by Hellman (1980), an algorithm gets blackbox access to a permutation $\pi:[N] \to [N]$ and takes as input a point $y \in [N]$, along with some auxiliary information about $\pi$. The algorithm must output $\pi^{1}(y)$.
The second problem we will discuss comes from communication complexity. There, a major goal is to prove a superlogarithmic lower bound for some problem for a superpolylogarithmic number of players in the numberontheforehead model. A classic candidate for such a lower bound is the pointer jumping problem, in which the goal of the players is to follow a sequence of pointers in a layered directed graph, where each player has one of the layers written on her forehead.
Although these two problems seem to have little in common, we will show that using classic cryptanalytic algorithms for permutation inversion we get new upper bounds for multiparty pointer jumping, improving the previous best bound of Pudl?k, R?dl, and Sgall (1997). As far as we know, this is the first application of a cryptanalytic algorithm to a communicationcomplexity problem.
Joint work with Henry CorriganGibbs.

Best,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
From rmhulett at stanford.edu Tue Nov 27 12:59:35 2018
From: rmhulett at stanford.edu (Reyna Marie Hulett)
Date: Tue, 27 Nov 2018 20:59:35 +0000
Subject: [theoryseminar] Theory Lunch  Dima Kogan
MessageID:
Hi folks,
In our penultimate theory lunch of the quarter, Dima Kogan will present "From Cryptanalysis to Communication Complexity"see abstract below. As always, please join us Thursday from noon to 1 pm in Gates 463A!

Abstract:
The first problem we will discuss comes from cryptanalysis. In the permutationinversion problem, first studied by Hellman (1980), an algorithm gets blackbox access to a permutation $\pi:[N] \to [N]$ and takes as input a point $y \in [N]$, along with some auxiliary information about $\pi$. The algorithm must output $\pi^{1}(y)$.
The second problem we will discuss comes from communication complexity. There, a major goal is to prove a superlogarithmic lower bound for some problem for a superpolylogarithmic number of players in the numberontheforehead model. A classic candidate for such a lower bound is the pointer jumping problem, in which the goal of the players is to follow a sequence of pointers in a layered directed graph, where each player has one of the layers written on her forehead.
Although these two problems seem to have little in common, we will show that using classic cryptanalytic algorithms for permutation inversion we get new upper bounds for multiparty pointer jumping, improving the previous best bound of Pudl?k, R?dl, and Sgall (1997). As far as we know, this is the first application of a cryptanalytic algorithm to a communicationcomplexity problem.
Joint work with Henry CorriganGibbs.

Best,
Reyna
 next part 
An HTML attachment was scrubbed...
URL:
From reingold at stanford.edu Tue Nov 27 15:01:21 2018
From: reingold at stanford.edu (Omer Reingold)
Date: Tue, 27 Nov 2018 15:01:21 0800
Subject: [theoryseminar] Fwd: Sexual harassment
InReplyTo:
References:
MessageID:
Hello everyone,
I am forwarding the email below, regarding the conduct of Yuval Peres,
partly as a fair public warning to our community. Another reason is to
emphasize again that sexual harassment and sexual misconduct go beyond the
narrow definitions of past generations, and for a good reason. As we and
others discussed repeatedly, sexual misconduct harms individual and hurts
the community (and thus hurts science as well). We can argue about the
specific rules to apply but the bottom line is that we should all be able
to share in the immense intellectual excitement that is TOC and the
closeness between collaborators that it brings without this intellectual
passion and closeness be misconstrued as anything else (in the spirit of
this rather old post
).
Finally, I want to remind everyone (and especially students), that you need
not be alone if you have been hurt (on campus, in a conference or elsewhere
as part of your work and even independent of work). All of the theory
faculty are here to support and there are multiple resources in CS and on
campus. Just contact whoever you feel most comfortable with.
Best wishes,
Omer
 Forwarded message 
From: Ehud Friedgut
Date: Tue, Nov 27, 2018 at 5:07 AM
Subject: Sexual harassment
To: ofer zeitouni , Itai Benjamini <
itai.benjamini at gmail.com>, Ronen Eldan , David Peleg <
david.peleg at weizmann.ac.il>, Michal Irani ,
Robert Krauthgamer , Jennifer Chayes <
jchayes at microsoft.com>, Christian Borgs ,
Anna Karlin , Ronitt Rubinfeld <
ronitt at csail.mit.edu>, Noga Alon , Benjamin Sudakov <
benny.sudakov at gmail.com>, Michael Krivelevich ,
Asaf Shapira , Michal Karonski ,
Andrzej Rucinski , Wojtek Samotij <
wojteksa at gmail.com>, Ron Peled , James R. Lee <
jrl at cs.washington.edu>, Nati Linial , Avi Wigderson <
avi at ias.edu>, Gil Kalai , Penny Haxell <
pehaxell at math.uwaterloo.ca>, Tamar Ziegler ,
Dorit Aharonov , , <
aaronson at cs.utexas.edu>, , , <
irani at ics.uci.edu>, , ,
,
Cc: Irit Dinur , Oded Goldreich <
oded.goldreich at weizmann.ac.il>
Hello all,
This is an email composed by Irit Dinur, Oded Goldreich and me.
The purpose of this email is to share with you concerns that we had
regarding the unethical behavior of Yuval Peres. The behavior we are
referring to includes several recent incidents from the past few years, on
top of the two ?big? cases of sexual harassment that led to severe
sanctions against him by his employer, Microsoft, and to the termination of
his connections with the University of Washington.
Together with two colleagues who are highly regarded and trusted by us, we
have first and secondhand testimonies (by people we trust without a shed
of doubt) of at least five additional cases of him approaching junior
female scientists, some of them students, with offers of intimate
nature, behavior that has caused its victims quite a bit of distress since
these offers were "insistent".
While the examples that we encountered from the last few years do not fall
under the category of sexual harassment from a legal point of view, they
certainly caused great discomfort to the victims, who were young female
scientists, putting them in a highly awkward situation, and creating an
atmosphere that they'd rather avoid (i.e., they would rather miss a
conference or a lecture than risk being subjected to repeated intimate
offers by him).
We wish to stress that his aggressive advances toward young women, usually
with no previous friendly connections with him, puts them in a vulnerable
position of fearing to cross a senior scientist who might have an impact on
their career, which is at a fragile stage. We believe that the questions of
whether or not Yuval Peres intended to make them uncomfortable, and whether
or not he would or could actually harm their scientific status are
irrelevant; the fact is that the victims felt very stressed to a point that
they'd rather miss professional events than risk encountering the same
situation again. Needless to say, it is the responsibility of senior
members of our community to avoid putting less senior members in such a
position.
Our current involvement with this issue was triggered by an invitation
Yuval Peres received to give a plenary talk at an international conference
next year. We felt that this invitation sends a highly undesirable message
to our community in general, and to the women he harassed in particular, as
if his transgressions are considered unimportant.
We sent an email conveying our concern to the organizers of the conference,
suggesting that they disinvite him. With our permission, they forwarded a
version of our letter (in which we made changes in order to protect the
identity of the women involved) to Yuval Peres. They did not reveal our
identity, rather they told him that this is a letter from "senior members
of the community". In our letter we included a paragraph describing a
general principle that should be followed. The principle is:
*A senior researcher should not approach a junior researcher with an
invitation that may be viewed as intimate or personal unless such an
invitation was issued in the past by this specific junior to that specific
senior. The point being that even if the senior researcher has no
intimate/personal intentions, such intentions may be read by the junior
researcher, placing the junior in an awkward situation and possibly causing
them great distress. Examples for such an invitation include any invitation
to a personal event in which only the senior and the junior will be present
(e.g., a twoperson dinner, a meeting in a private home, etc).*
Yuval's reply was rather laconic, in particular, he did not address the
issue of his behavior in the past couple of years. However, he did write:
"I certainly embrace the principle described in boldface in the letter.
This seems to be the right approach for any senior scientist these days."
The reason we are copying this to all of you (as opposed, for example, to
using bcc) is related to the i
slanders
'
paradox
:
we believe that the fact that everyone knows that everyone knows is a
significant boost to holding Yuval Peres accountable for his future
actions. We're also bcc'ing several young women who already aware of Yuval
Peres's actions, in order to keep them in the know too.
We understand that sending this out to a large number of people without
offering Yuval Peres the chance to respond may be considered unfair.
However, after weighing the pros and cons carefully we believe this is a
good course of action. First of all, because it is clear that the victims
did not invent his offers and their ensuing feelings of anxiety and stress.
Secondly, we know that Yuval Peres has been confronted in a face to face
conversation by a senior colleague, and it did not end his behavior, so we
think it's important to stay vigilant in protecting the younger members of
our community. Thirdly, the information in this letter will reach (or has
already reached) almost all of you in any case, so the main effect of the
letter is making what everyone knows into public knowledge. Finally,
although his response to the organizers did include the minimum of
declaring he accepts the guiding principle that we stated, it did not
include any reference to the ongoing behavior we described  neither regret
nor concern nor denial. So it's not easy to assume that he truly intends to
mend his ways.
We hope that our actions will contribute to the future of our community as
an environment that offers all a pleasant and nonthreatening atmosphere.
Sincerely,
Irit Dinur, Ehud Friedgut, Oded Goldreich
 next part 
An HTML attachment was scrubbed...
URL:
From reingold at stanford.edu Wed Nov 28 06:45:14 2018
From: reingold at stanford.edu (Omer Reingold)
Date: Wed, 28 Nov 2018 06:45:14 0800
Subject: [theoryseminar] Sexual harassment
InReplyTo:
References:
MessageID:
Dear all,
In the name of fairness, I am including below a response from Yuval Peres
to the email I forwarded before. I suggest that anybody who shared my own
email would do the same. As you can imagine, there is a considerable amount
of discussion on the original email list. I will not share any additional
parts of this discussion or any additional back and forth with Yuval as my
point goes way beyond his case. I am hoping that my emails are not flooding
uncomfortable or painful memories to any of you (as it is my understanding
that experiences of harrastment are much more common than we would like to
believe). Let me conclude by reiterating that each and every one of the
theory faculty as well as a multitude of resources at Stanford are
committed to your support and to a safe and welcoming work environment.
Take care,
Omer
>From Yuval Peres: 
Dear Ehud, Irit and Oded, Dear email recipients:
This is my community, and I greatly appreciate each and every one of you.
As you can imagine, I am sad to see this email but understand it arises
from genuine concern of the authors.
(I do wish they had written to me directly rather than through
intermediaries).
I disagree with some claims in the email but let me start by emphasizing
that I embrace the passage:
*A senior researcher should not approach a junior researcher with an
invitation that may be viewed as intimate or personal unless such an
invitation was issued in the past by this specific junior to that specific
senior. The point being that even if the senior researcher has no
intimate/personal intentions, such intentions may be read by the junior
researcher, placing the junior in an awkward situation and possibly causing
them great distress. Examples for such an invitation include any invitation
to a personal event in which only the senior and the junior will be present
(e.g., a twoperson dinner, a meeting in a private home, etc).*
*I regret all cases in the past where I have not followed this principle.* I
had no intention to harass anyone but must have been tone deaf not to
recognize that I was making some people very uncomfortable. As I wrote
above, I promise to adhere to this principle in the future. In order to
show that I am not trivializing the concerns of Ehud, Irit and Oded*, I
will regretfully rescind my acceptance to deliver a keynote talk in the
conference that sparked their message*. I apologize to the organizers of
that conference and wrote to them separately.
I have collaborated with about 300 researchers in my career, many of them
women, and most of my success is due to those collaborations. (A lot of
those collaborations took place in twoperson lunches and dinners but
usually when the collaborators knew each other well). In particular, I
consider my early work with Claire Mathieu and my recent book with Anna
Karlin as some of the high points of my career. In the last SODA/ANALCO I
gave four talks, three of them based on collaborations with brilliant
young women. I completely agree that it is crucial that we present a
welcoming atmosphere to these women.
If any one of you has further suggestions, you are very welcome to write
to me personally at this address.
Yours,
Yuval

On Tue, Nov 27, 2018, 3:01 PM Omer Reingold Hello everyone,
>
> I am forwarding the email below, regarding the conduct of Yuval Peres,
> partly as a fair public warning to our community. Another reason is to
> emphasize again that sexual harassment and sexual misconduct go beyond the
> narrow definitions of past generations, and for a good reason. As we and
> others discussed repeatedly, sexual misconduct harms individual and hurts
> the community (and thus hurts science as well). We can argue about the
> specific rules to apply but the bottom line is that we should all be able
> to share in the immense intellectual excitement that is TOC and the
> closeness between collaborators that it brings without this intellectual
> passion and closeness be misconstrued as anything else (in the spirit of
> this rather old post
> ).
> Finally, I want to remind everyone (and especially students), that you need
> not be alone if you have been hurt (on campus, in a conference or elsewhere
> as part of your work and even independent of work). All of the theory
> faculty are here to support and there are multiple resources in CS and on
> campus. Just contact whoever you feel most comfortable with.
>
> Best wishes,
> Omer
>
>
>  Forwarded message 
> From: Ehud Friedgut
> Date: Tue, Nov 27, 2018 at 5:07 AM
> Subject: Sexual harassment
> To: ofer zeitouni , Itai Benjamini <
> itai.benjamini at gmail.com>, Ronen Eldan , David
> Peleg , Michal Irani <
> michal.irani at weizmann.ac.il>, Robert Krauthgamer <
> robert.krauthgamer at weizmann.ac.il>, Jennifer Chayes ,
> Christian Borgs , Anna Karlin <
> karlin at cs.washington.edu>, Ronitt Rubinfeld , Noga
> Alon , Benjamin Sudakov ,
> Michael Krivelevich , Asaf Shapira <
> asafico at post.tau.ac.il>, Michal Karonski , Andrzej
> Rucinski , Wojtek Samotij ,
> Ron Peled , James R. Lee ,
> Nati Linial , Avi Wigderson , Gil Kalai <
> gil.kalai at gmail.com>, Penny Haxell , Tamar
> Ziegler , Dorit Aharonov <
> doria at cs.huji.ac.il>, , , <
> reingold at stanford.edu>, , , <
> shafi.goldwasser at weizmann.ac.il>, , <
> salil_vadhan at harvard.edu>,
> Cc: Irit Dinur , Oded Goldreich <
> oded.goldreich at weizmann.ac.il>
>
>
> Hello all,
>
> This is an email composed by Irit Dinur, Oded Goldreich and me.
> The purpose of this email is to share with you concerns that we had
> regarding the unethical behavior of Yuval Peres. The behavior we are
> referring to includes several recent incidents from the past few years, on
> top of the two ?big? cases of sexual harassment that led to severe
> sanctions against him by his employer, Microsoft, and to the termination of
> his connections with the University of Washington.
>
> Together with two colleagues who are highly regarded and trusted by us, we
> have first and secondhand testimonies (by people we trust without a shed
> of doubt) of at least five additional cases of him approaching junior
> female scientists, some of them students, with offers of intimate
> nature, behavior that has caused its victims quite a bit of distress since
> these offers were "insistent".
>
> While the examples that we encountered from the last few years do not fall
> under the category of sexual harassment from a legal point of view, they
> certainly caused great discomfort to the victims, who were young female
> scientists, putting them in a highly awkward situation, and creating an
> atmosphere that they'd rather avoid (i.e., they would rather miss a
> conference or a lecture than risk being subjected to repeated intimate
> offers by him).
>
> We wish to stress that his aggressive advances toward young women, usually
> with no previous friendly connections with him, puts them in a vulnerable
> position of fearing to cross a senior scientist who might have an impact on
> their career, which is at a fragile stage. We believe that the questions of
> whether or not Yuval Peres intended to make them uncomfortable, and whether
> or not he would or could actually harm their scientific status are
> irrelevant; the fact is that the victims felt very stressed to a point that
> they'd rather miss professional events than risk encountering the same
> situation again. Needless to say, it is the responsibility of senior
> members of our community to avoid putting less senior members in such a
> position.
>
>
>
> Our current involvement with this issue was triggered by an invitation
> Yuval Peres received to give a plenary talk at an international conference
> next year. We felt that this invitation sends a highly undesirable message
> to our community in general, and to the women he harassed in particular, as
> if his transgressions are considered unimportant.
>
> We sent an email conveying our concern to the organizers of the
> conference, suggesting that they disinvite him. With our permission, they
> forwarded a version of our letter (in which we made changes in order to
> protect the identity of the women involved) to Yuval Peres. They did not
> reveal our identity, rather they told him that this is a letter from
> "senior members of the community". In our letter we included a paragraph
> describing a general principle that should be followed. The principle is:
>
>
>
> *A senior researcher should not approach a junior researcher with an
> invitation that may be viewed as intimate or personal unless such an
> invitation was issued in the past by this specific junior to that specific
> senior. The point being that even if the senior researcher has no
> intimate/personal intentions, such intentions may be read by the junior
> researcher, placing the junior in an awkward situation and possibly causing
> them great distress. Examples for such an invitation include any invitation
> to a personal event in which only the senior and the junior will be present
> (e.g., a twoperson dinner, a meeting in a private home, etc).*
>
>
>
> Yuval's reply was rather laconic, in particular, he did not address the
> issue of his behavior in the past couple of years. However, he did write:
>
>
> "I certainly embrace the principle described in boldface in the letter.
>
> This seems to be the right approach for any senior scientist these days."
>
>
>
>
>
> The reason we are copying this to all of you (as opposed, for example, to
> using bcc) is related to the i
>
> slanders
> '
> paradox
> :
> we believe that the fact that everyone knows that everyone knows is a
> significant boost to holding Yuval Peres accountable for his future
> actions. We're also bcc'ing several young women who already aware of Yuval
> Peres's actions, in order to keep them in the know too.
>
>
>
>
>
> We understand that sending this out to a large number of people without
> offering Yuval Peres the chance to respond may be considered unfair.
> However, after weighing the pros and cons carefully we believe this is a
> good course of action. First of all, because it is clear that the victims
> did not invent his offers and their ensuing feelings of anxiety and stress.
> Secondly, we know that Yuval Peres has been confronted in a face to face
> conversation by a senior colleague, and it did not end his behavior, so we
> think it's important to stay vigilant in protecting the younger members of
> our community. Thirdly, the information in this letter will reach (or has
> already reached) almost all of you in any case, so the main effect of the
> letter is making what everyone knows into public knowledge. Finally,
> although his response to the organizers did include the minimum of
> declaring he accepts the guiding principle that we stated, it did not
> include any reference to the ongoing behavior we described  neither regret
> nor concern nor denial. So it's not easy to assume that he truly intends to
> mend his ways.
>
>
>
> We hope that our actions will contribute to the future of our community as
> an environment that offers all a pleasant and nonthreatening atmosphere.
>
>
> Sincerely,
>
> Irit Dinur, Ehud Friedgut, Oded Goldreich
>
 next part 
An HTML attachment was scrubbed...
URL:
From ofirgeri at stanford.edu Wed Nov 28 09:25:34 2018
From: ofirgeri at stanford.edu (Ofir Geri)
Date: Wed, 28 Nov 2018 17:25:34 +0000
Subject: [theoryseminar] Theory Seminar (11/28): Urmila Mahadev
InReplyTo:
References:
MessageID:
Reminder: Urmila Mahadev's talk is today at 3pm in Gates 463A.
________________________________
From: Ofir Geri
Sent: Sunday, November 25, 2018 11:24:46 PM
To: thseminar at cs.stanford.edu
Subject: Theory Seminar (11/28): Urmila Mahadev
Hi all,
This Wednesday, Urmila Mahadev (UC Berkeley) will be giving a theory seminar talk on Classical Verification of Quantum Computations (abstract below). The talk will be at 3:00 PM in Gates 463A.
Please note that this week's talk is on Wednesday (and there will be no theory seminar on Friday).
Hope to see you there!
Ofir
Classical Verification of Quantum Computations
Speaker: Urmila Mahadev (UC Berkeley)
We present the first protocol allowing a classical computer to interactively verify the result of an efficient quantum computation. We achieve this by constructing a measurement protocol, which allows a classical string to serve as a commitment to a quantum state. The protocol forces the prover to behave as follows: the prover must construct an n qubit state of his choice, nonadaptively measure each qubit in the Hadamard or standard basis as directed by the verifier, and report the measurement results to the verifier. The soundness of this protocol is enforced based on the assumption that the learning with errors problem is computationally intractable for efficient quantum machines.
 next part 
An HTML attachment was scrubbed...
URL: