TCS+

A carbon-free dissemination of ideas across the globe.

Guest post: TCS for All (previously TCS Women) Spotlight Workshop at STOC 2024/Theory Fest: Travel grants and call for speaker nominations.

Guest post by the organizers of TCS for All (previously TCS Women).

You are cordially invited to our TCS for All Spotlight Workshop! The workshop will be held on June 24, 2024, at 12:30 pm – 2 pm, and 4:15 pm – 5:45 pm in Vancouver, Canada as part of the 56th Symposium on Theory of Computing (STOC) and TheoryFest! The workshop is open to all.

More information about the workshop would be made available here: https://sigact.org/tcsforall/. In particular, we would like to highlight the TCS for All Travel Scholarships (deadline April 28th) and a call for nominations for Rising Stars talks at the workshop (deadline April 28th).  More information on those are below.

Hope to see you in Vancouver!

TCS for All Travel Scholarship:

TCS for All Travel Scholarships are intended for researchers at the beginning of their career. This scholarship is being made available for minorities in TCS, and anyone who identifies as such is welcome to apply; this scholarship is open to both US and international students. Preference will be given to students at the beginning of their studies. If we have sufficient funding, we will give awards to more senior students and possibly even postdocs.

To apply, you will need to fill out the following form by April 28th, 2024 (11:59 pm PDT) in which you provide basic information about yourself, an estimate of your expenses, and a brief statement:

Apply for a travel grant here.

In addition, you will need to have your advisor (or department head or other faculty mentor if you do not yet have an advisor) send a letter of support to tcswomen@gmail.com by April 28th. Your advisor’s letter should also describe the availability of other travel funds.  Note for advisors: Specifics about alternative funding are very helpful.  Statements like “funding is tight” are not very helpful. This letter should be sent with the subject line “support letter for [your name]”. This is very important. Your application is not complete without this letter.

Late applications (after April 28th) will not be accepted. You will be notified about your status by May 5th, which is prior to the STOC early registration deadline and hotel cut-off deadline.

Notes: Receipts will be required for all travel awards, and reimbursements will be made after the conference. Food or visa expenses will not be reimbursed.

Nominations for Rising Star talks:

We invite nominations for speakers in our Rising Star talks at the TCS for All Spotlight Workshop at STOC 2024. To be eligible, your nominee has to be a senior PhD student with expected graduation no later than August 2024, or a postdoc in theoretical computer science (all topics represented at STOC are welcome), an underrepresented minority, and not a speaker at a previous TCS Women Spotlight Workshop.  Preference will be given to speakers who are currently on the job market for postdoctoral/faculty positions, or who expect to be on the job market in Fall 2024.

You can make your nomination by filling this form by April 28th: https://forms.gle/WAZsXRm2wuxhdukK7

TCS+ talk: Wednesday, April 10 — Zeyong Li, National University of Singapore (NUS)

The next TCS+ talk will take place this coming Wednesday, April 10th at 10:30 AM Eastern Time (7:30 AM Pacific Time, 16:30 Central Summer European Time, 14:30 UTC: check yours here). Zeyong Li from National University of Singapore (NUS) will speak about “Simple Circuit Lower Bound via Algorithm for the Range Avoidance Problem” (abstract below). Note the slightly unusual time!

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: We present a simple single-valued FS_2P algorithm for the Range Avoidance Problem. As a result, we obtain the circuit lower bound S_2E \not \subset i.o.-SIZE[2^n/n] and many other corollaries:

  1. Almost-everywhere near-maximum circuit lower bound for Sigma_2E \intersect Pi_2E and for ZPE^NP.
  2. Pseudodeterministic FZPP^NP constructions for: Ramsey graphs, rigid matrices, pseudorandom generators, two-source extractors, linear codes, hard truth tables, and K^poly-random strings.

This talk is based on the paper titled ‘Symmetric Exponential Time Requires Near-Maximum Circuit Size: Simplified, Truly Uniform’, to appear in STOC 2024.

TCS+ talk: Wednesday, March 20 — Huan Li, University of Pennsylvania

TCS+ is back for a new season! The next TCS+ talk will take place this coming Wednesday, March 20th at 1:00 PM Eastern Time (10:00 AM Pacific Time, 18:00 Central European Time, 17:00 UTC). Huan Li from University of Pennsylvania will speak about “Parallel Approximate Maximum Flows in Near-Linear Work and Polylogarithmic Depth” (abstract below).

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: We present a parallel algorithm for the (1-\epsilon)-approximate maximum flow problem in capacitated, undirected graphs with n vertices and $m$ edges, achieving O(\epsilon^{-3}\text{polylog} n) depth and O(m \epsilon^{-3} \text{polylog} n) work in the PRAM model. Although near-linear time sequential algorithms for this problem have been known for almost a decade, no parallel algorithms that simultaneously achieved polylogarithmic depth and near-linear work were known. At the heart of our result is a polylogarithmic depth, near-linear work recursive algorithm for computing congestion approximators. Our algorithm involves a recursive step to obtain a low-quality congestion approximator followed by a “boosting” step to improve its quality which prevents a multiplicative blow-up in error. Similar to Peng [SODA’16], our boosting step builds upon the hierarchical decomposition scheme of R\”acke, Shah, and T\”aubig [SODA’14]. A direct implementation of this approach, however, leads only to an algorithm with n^{o(1)} depth and m^{1+o(1)} work. To get around this, we introduce a new hierarchical decomposition scheme, in which we only need to solve maximum flows on subgraphs obtained by contracting vertices, as opposed to vertex-induced subgraphs used in R\”acke, Shah, and T\”aubig [SODA’14]. In particular, we are able to directly extract congestion approximators for the subgraphs from a congestion approximator for the entire graph, thereby avoiding additional recursion on those subgraphs. Along the way, we also develop a parallel flow-decomposition algorithm that is crucial to achieving polylogarithmic depth and may be of independent interest.

TCS+ talk: Wednesday, December 13 — Aaron Bernstein, Rutgers University

The next TCS+ talk (and the last of 2023!) will take place this coming Wednesday, December 13th at 1:00 PM Eastern Time (10:00 AM Pacific Time, 19:00 Central European Time, 18:00 UTC). Aaron Bernstein from Rutgers University will speak about “Negative-Weight Single-Source Shortest Paths in Near-linear Time” (abstract below).

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: We present a randomized algorithm that computes single-source shortest paths (SSSP) in O(m\log^8(n)\log W) time when edge weights are integral and can be negative. This essentially resolves the classic negative-weight SSSP problem. The previous bounds are \tilde O((m+n^{1.5})\log W) [BLNPSSSW FOCS’20] and m^{4/3+o(1)}\log W [AMV FOCS’20]. Near-linear time algorithms were known previously only for the special case of planar directed graphs [Fakcharoenphol and Rao FOCS’01].

In contrast to all recent developments that rely on sophisticated continuous optimization methods and dynamic algorithms, our algorithm is simple: it requires only a simple graph decomposition and elementary combinatorial tools. In fact, ours is the first combinatorial algorithm for negative-weight SSSP to break through the classic \tilde O(m\sqrt{n}\log W) bound from over three decades ago [Gabow and Tarjan SICOMP’89].

Paper: https://arxiv.org/abs/2203.03456

TCS+ talk: Wednesday, November 29 — Kasper Green Larsen, Aarhus University

The next TCS+ talk will take place this coming Wednesday, November 29th at 10:30 AM Eastern Time (7:30 AM Pacific Time, 16:30 Central European Time, 15:30 UTC — note the unusual time!). Kasper Green Larsen from Aarhus University will speak about “Bagging is an Optimal PAC Learner” (abstract below).

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: Determining the optimal sample complexity of PAC learning in the realizable setting was a central open problem in learning theory for decades. Finally, the seminal work by Hanneke (2016) gave an algorithm with a provably optimal sample complexity. His algorithm is based on a careful and structured sub-sampling of the training data and then returning a majority vote among hypotheses trained on each of the sub-samples. While being a very exciting theoretical result, it has not had much impact in practice, in part due to inefficiency, since it constructs a polynomial number of sub-samples of the training data, each of linear size. I

n this talk, we prove the surprising result that the practical and classic heuristic Bagging (a.k.a. bootstrap aggregation), due to Breiman (1996), is in fact also an optimal PAC learner. Bagging pre-dates Hanneke’s algorithm by twenty years and is taught in most undergraduate machine learning courses. Moreover, we show that it only requires a logarithmic number of sub-samples to reach optimality.

This work was published at COLT’23 and received the Best Paper Award.

TCS+ talk: Wednesday, November 15 — Palak Jain and Satchit Sivakumar, Boston University

The next TCS+ talk will take place this coming Wednesday, November 15th at 1:00 PM Eastern Time (10:00 AM Pacific Time, 19:00 Central European Time, 18:00 UTC). Palak Jain and Satchit Sivakumar from Boston University will speak about “The Price of Differential Privacy under Continual Observation” (abstract below).

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: This talk will be on the accuracy of differentially private mechanisms in the continual release model. A continual release mechanism receives a sensitive dataset as a stream of T inputs and produces, after receiving each input, an output that is accurate for all the inputs received so far.

We describe a ‘sequential embedding’ framework to prove lower bounds for the continual release model via reductions from batch model problems. (In the batch model, the algorithm receives the data as one batch and produces a single output.) We show how this framework can be used to get strong lower bounds for some fundamental problems that are widely studied; these are the first strong lower bounds on the error of continual release mechanisms and witness a polynomial in T separation between the batch model and continual release model. Previous work shows only a logarithmic in T gap between the worst-case error achievable in these two models.

Our lower bounds assume only that privacy holds for streams fixed in advance (the ‘nonadaptive’ setting). We also discuss a model that allows for adaptively selected inputs. It captures dependencies that arise in many applications of continual release. In general, both privacy and accuracy are harder to attain in this model. Nevertheless, we analyze several algorithms in the new model and, in particular, show that some of our lower bounds are matched by the error of simple algorithms whose privacy holds even for adaptively selected streams.

This talk is based on works with Iden Kalemaj, Sofya Raskhodnikova, and Adam Smith (abs/2112.00828 and abs/2306.06723).

TCS+ talk: Wednesday, November 1 — Victor Reis, University of Washington

The next TCS+ talk will take place this coming Wednesday, November 1st at 1:00 PM Eastern Time (10:00 AM Pacific Time, 18:00 Central European Time, 17:00 UTC). Victor Reis from University of Washington will speak about “The Subspace Flatness Conjecture and Faster Integer Programming” (abstract below).

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: In a seminal paper, Kannan and Lovász (1988) considered a quantity \mu_{KL}(\Lambda,K) which denotes the best volume-based lower bound on the covering radius \mu(\Lambda,K) of a convex body K with respect to a lattice \Lambda. Kannan and Lovász proved that \mu(\Lambda,K) \leq n \cdot \mu_{KL}(\Lambda,K) and the Subspace Flatness Conjecture by Dadush (2012) claims a O(\log n) factor suffices, which would match the lower bound from the work of Kannan and Lovász.

We settle this conjecture up to a constant in the exponent by proving that \mu(\Lambda,K) \leq O(log^3(n)) \cdot \mu_{KL} (\Lambda,K). Our proof is based on the Reverse Minkowski Theorem due to Regev and Stephens-Davidowitz (2017). Following the work of Dadush (2012, 2019), we obtain a (log n)^{4n}-time randomized algorithm to solve integer programs in n variables. Another implication of our main result is a near-optimal flatness constant of O(n log^3(n)). Joint work with Thomas Rothvoss.

TCS+ talk: Wednesday, September 27 — Hanlin Ren, University of Oxford

It’s September, and we are back!

The next TCS+ talk, first of our brand new season of TCS+, will take place next Wednesday, September 27th at 1:00 PM Eastern Time (10:00 AM Pacific Time, 19:00 Central European Time, 17:00 UTC (check your timezone here)). Hanlin Ren from University of Oxford will speak about “Polynomial-Time Pseudodeterministic Construction of Primes” (abstract below).

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: A randomized algorithm for a search problem is *pseudodeterministic* if it produces a fixed canonical solution to the search problem with high probability. In their seminal work on the topic, Gat and Goldwasser posed as their main open problem whether prime numbers can be pseudodeterministically constructed in polynomial time.

We provide a positive solution to this question in the infinitely-often regime. In more detail, we give an *unconditional* polynomial-time randomized algorithm B such that, for infinitely many values of n, B(1^n) outputs a canonical n-bit prime p_n with high probability. More generally, we prove that for every dense property Q of strings that can be decided in polynomial time, there is an infinitely-often pseudodeterministic polynomial-time construction of strings satisfying Q. This improves upon a subexponential-time construction of Oliveira and Santhanam.

Our construction uses several new ideas, including a novel bootstrapping technique for pseudodeterministic constructions, and a quantitative optimization of the uniform hardness-randomness framework of Chen and Tell, using a variant of the Shaltiel—Umans generator.

Guest post: STOCial activities, Graduating Bits, and Junior/Senior Lunch at STOC 2023

If you are attending STOC’23 as part of the TheoryFest next week, don’t forget to have a look at the STOCial Program!

GIF of a man with a beret saying "It is fun!"

In particular, and to get to the point of this announcement: if you are graduating or will be on the job market soon, consider participating to the Graduating Bits Session on Wednesday: 1-2 slides, 2 minutes, entirely yours to present and pitch your research to the STOC attendees! To register, fill this form before Tuesday evening (up to 25 slots, first-come-first-served).

Following the by-now-well-established tradition, there will also be a “senior/junior lunch” on Thursday, during the 12:30-2pm time slot: put your name here as a “senior” (broadly construed) or a “junior” academic, for informal discussions, academic advice, and general questions over lunch.

See you next week at the TheoryFest!

Clément, on behalf of the STOCial Committee (Barna Saha, Clément Canonne, Elena Grigorescu, and Raghu Meka)

TCS+ talk: Wednesday, May 31 — Paul Gölz, Harvard University

The next TCS+ talk will take place this coming Wednesday, May 31th at 1:00 PM Eastern Time (10:00 AM Pacific Time, 19:00 Central European Time, 17:00 UTC). Paul Gölz from Harvard University will speak about “News from Algorithmic Democracy: Proportional Representation for Preferences and Demographics” (abstract below).

You can reserve a spot as an individual or a group to join us live by signing up on the online form. Registration is not required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The recorded talk will also be posted on our website afterwards) As usual, for more information about the TCS+ online seminar series and the upcoming talks, or to suggest a possible topic or speaker, please see the website.

Abstract: How do you fill a knapsack with city projects to fund, in a way that aligns with voters’ preferences? And how do you choose a committee that represents the population’s makeup in terms of gender, age, etc.? Both questions are central to experiments with new forms of democracy in practice, and both have spurred an exploration for the right algorithms in computational social choice. In this talk, I will survey advances along these thrusts: proposed algorithms, guarantees offered and sought after, as well as technical challenges and connections.