BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//הפקולטה למדעי המחשב והמידע - ECPv6.15.15//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-ORIGINAL-URL:https://cis.haifa.ac.il
X-WR-CALDESC:Events for הפקולטה למדעי המחשב והמידע
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Asia/Jerusalem
BEGIN:DAYLIGHT
TZOFFSETFROM:+0200
TZOFFSETTO:+0300
TZNAME:IDT
DTSTART:20250328T000000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0300
TZOFFSETTO:+0200
TZNAME:IST
DTSTART:20251025T230000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0200
TZOFFSETTO:+0300
TZNAME:IDT
DTSTART:20260327T000000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0300
TZOFFSETTO:+0200
TZNAME:IST
DTSTART:20261024T230000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0200
TZOFFSETTO:+0300
TZNAME:IDT
DTSTART:20270326T000000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0300
TZOFFSETTO:+0200
TZNAME:IST
DTSTART:20271030T230000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;VALUE=DATE:20260331
DTEND;VALUE=DATE:20260409
DTSTAMP:20260416T022323
CREATED:20260315T071432Z
LAST-MODIFIED:20260315T071829Z
UID:6485-1774915200-1775692799@cis.haifa.ac.il
SUMMARY:חופשת פסח 31.3.26-8.4.26
DESCRIPTION:חופשת פסח: מיום שלישי\, י"ג ניסן\, תשפ"ו\, 29.03.26\, עד יום רביעי\, כ"א ניסן\, תשפ"ו\, 08.04.26. חזרה ללימודים ביום חמישי\, כ"ב ניסן\, תשפ"ו\, 09.04.26.
URL:https://cis.haifa.ac.il/event/%d7%97%d7%95%d7%a4%d7%a9%d7%aa-%d7%a4%d7%a1%d7%97/
CATEGORIES:אירועים מדעי המחשב,אירועים מערכות מידע
END:VEVENT
BEGIN:VEVENT
DTSTART;VALUE=DATE:20260322
DTEND;VALUE=DATE:20260323
DTSTAMP:20260416T022323
CREATED:20260226T122908Z
LAST-MODIFIED:20260309T133921Z
UID:6410-1774137600-1774223999@cis.haifa.ac.il
SUMMARY:22.3.26 פתיחת סמסטר ב'
DESCRIPTION:פתיחת סמסטר ב' באוניברסיטת חיפה לשנת הלימודים תשפ"ו צפויה להתקיים ביום ראשון\, 22 במרץ 2026. עבור סטודנטים חדשים המעוניינים להצטרף למסלולי הלימוד בסמסטר אביב זה\, ההרשמה נפתחה כבר ב-30 בנובמבר 2025. הסמסטר\, המביא עמו רוח של התחדשות לקמפוס העיר\, יימשך עד לסיומו ב-23 ביוני 2026\, וכולל בתוכו את חופשת הפסח המסורתית וימי זיכרון ומועד לאומיים.
URL:https://cis.haifa.ac.il/event/22-3-26-%d7%a4%d7%aa%d7%99%d7%97%d7%aa-%d7%a1%d7%9e%d7%a1%d7%98%d7%a8-%d7%91/
CATEGORIES:אירועים מדעי המחשב,אירועים מערכות מידע,סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260312T110000
DTEND;TZID=Asia/Jerusalem:20260312T120000
DTSTAMP:20260416T022323
CREATED:20260219T134325Z
LAST-MODIFIED:20260219T135412Z
UID:6321-1773313200-1773316800@cis.haifa.ac.il
SUMMARY:From Attacks to Security-Enhancing Insights in NLP Models
DESCRIPTION:Abstract: \nRecent advances in natural language processing (NLP) have given rise to transformative models\, including large language models (LLMs) and text retrievers. Still\, critical concerns remain regarding the security of these models: chiefly\, LLMs can be jailbroken and misused (e.g.\, to launch cyberattacks)\, and text retrievers in search applications can be manipulated to prioritize adversary-chosen content. In this talk\, I will present our recent efforts toward making LLMs and text retrievers more secure. In particular\, I will show how potent attacks can provide explanations for models' vulnerabilities\, which\, in turn\, enable us to enhance security. Crucially\, I will also demonstrate how our insights can inform the design of even stronger attacks\, establishing a cycle that guides continuous model improvements. \nBased on joint work with Matan Ben-Tov and Mor Geva. \nBio:  \nMahmood Sharif is a senior lecturer at the Blavatnik School of Computer Science at Tel Aviv University\, where he directs the privacy\, learning\, usability\, and security (PLUS) group—a research group primarily working at the intersections of computer security and privacy with machine learning\, specifically adversarial machine learning\, and with human factors. Mahmood obtained his Ph.D. from Carnegie Mellon University\, where he was affiliated with the CyLab Security and Privacy Institute. Before joining Tel Aviv University\, Mahmood was a postdoctoral researcher in the VMware Research Group and a principal research engineer in the NortonLifeLock Research Group. His work has been recognized by multiple awards\, including an Intel Rising Star Faculty award and a Maof prize for outstanding new faculty.
URL:https://cis.haifa.ac.il/event/from-attacks-to-security-enhancing-insights-in-nlp-models/
LOCATION:ביניין עמיר\, הנמל 65\, חדר 507\, חיפה\, North\, Israel
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;VALUE=DATE:20260302
DTEND;VALUE=DATE:20260303
DTSTAMP:20260416T022323
CREATED:20260226T123353Z
LAST-MODIFIED:20260226T123353Z
UID:6412-1772409600-1772495999@cis.haifa.ac.il
SUMMARY:יום פתוח למתעניינים בלימודים בתכנית "אתגר"
DESCRIPTION:מוזמנים ליום פתוח לתכנית "אתגר" בפקולטה למדעי המחשב והמידע\, אוניברסיטת חיפה! \nבואו להכיר מקרוב את מסלול הלימודים הייחודי\, לשמוע על אפשרויות הקבלה\, לפגוש את צוות התכנית ולקבל מענה אישי לכל השאלות. \n📅 02 במרץזו ההזדמנות שלכם לעשות צעד ראשון לעבר עתיד אקדמי משמעותי. נשמח לראותכם!
URL:https://cis.haifa.ac.il/event/%d7%99%d7%95%d7%9d-%d7%a4%d7%aa%d7%95%d7%97-%d7%9c%d7%9e%d7%aa%d7%a2%d7%a0%d7%99%d7%99%d7%a0%d7%99%d7%9d-%d7%91%d7%9c%d7%99%d7%9e%d7%95%d7%93%d7%99%d7%9d-%d7%91%d7%aa%d7%9b%d7%a0%d7%99%d7%aa-%d7%90/
CATEGORIES:אירועים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;VALUE=DATE:20260226
DTEND;VALUE=DATE:20260227
DTSTAMP:20260416T022323
CREATED:20251216T095515Z
LAST-MODIFIED:20251216T095515Z
UID:3550-1772064000-1772150399@cis.haifa.ac.il
SUMMARY:יום עיון של האגודה הישראלית לסינטומטריקה
DESCRIPTION:יום עיון של האגודה הישראלית לסינטומטריקה בנושא "דברים שרציתם לדעת ואמרו לכם לא לשאול" עם הרצאות וסדנאות פרקטיות בתחום הסינטומטריקה בדגש על שינויים אחרונים בארץ ובעולם.
URL:https://cis.haifa.ac.il/event/%d7%99%d7%95%d7%9d-%d7%a2%d7%99%d7%95%d7%9f-%d7%a9%d7%9c-%d7%94%d7%90%d7%92%d7%95%d7%93%d7%94-%d7%94%d7%99%d7%a9%d7%a8%d7%90%d7%9c%d7%99%d7%aa-%d7%9c%d7%a1%d7%99%d7%a0%d7%98%d7%95%d7%9e%d7%98%d7%a8/
LOCATION:ביניין עמיר\, הנמל 65\, חדר 507\, חיפה\, North\, Israel
CATEGORIES:אירועים מערכות מידע
ORGANIZER;CN="%D7%93'%D7%A8 %D7%98%D7%93%D7%99 %D7%9C%D7%96%D7%91%D7%A0%D7%99%D7%A7":MAILTO:lazebnik.teddy@is.haifa.ac.il
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260205T130000
DTEND;TZID=Asia/Jerusalem:20260205T140000
DTSTAMP:20260416T022323
CREATED:20260127T063157Z
LAST-MODIFIED:20260127T063157Z
UID:5965-1770296400-1770300000@cis.haifa.ac.il
SUMMARY:Nataly Brukhim (IAS and DIMACS) - via ZOOM
DESCRIPTION:
URL:https://cis.haifa.ac.il/event/nataly-brukhim-ias-and-dimacs-via-zoom/
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260129T110000
DTEND;TZID=Asia/Jerusalem:20260129T120000
DTSTAMP:20260416T022323
CREATED:20260127T063054Z
LAST-MODIFIED:20260127T063054Z
UID:5963-1769684400-1769688000@cis.haifa.ac.il
SUMMARY:Noam Razin (Princeton University) - Fundamentals of Aligning General-Purpose AI
DESCRIPTION:Abstract:\nThe field of artificial intelligence (AI) is undergoing a paradigm shift\, moving from neural\nnetworks trained for narrowly defined tasks (e.g.\, image classification and machine translation)\nto general-purpose models such as ChatGPT. These models are trained at unprecedented\nscales to perform a wide range of tasks\, from providing travel recommendations to solving\nOlympiad-level math problems. As they are increasingly adopted in society\, a central challenge\nis to ensure the alignment of general-purpose models with human preferences. In this talk\, I will\npresent a series of works that reveal fundamental pitfalls in existing alignment methods. In\nparticular\, I will show that they can: (1) suffer from a flat objective landscape that hinders\noptimization\, and (2) fail to reliably increase the likelihood of generating preferred outputs\,\nsometimes even causing the model to generate outputs with an opposite meaning. Beyond\ncharacterizing these pitfalls\, our theory provides quantitative measures for identifying when they\noccur\, suggests preventative guidelines\, and has led to the development of new data selection\nand alignment algorithms\, validated at large scale in real-world settings. Our contributions\naddress both efficiency challenges and safety risks that may arise in the alignment process. I\nwill conclude with an outlook on future directions\, toward building a practical theory in the age of\ngeneral-purpose AI. \nShort bio:\nNoam Razin is a Postdoctoral Fellow at Princeton Language and Intelligence\, Princeton\nUniversity. His research focuses on the fundamentals of artificial intelligence (AI). By combining mathematical analyses with systematic experimentation\, he aims to develop theories that shed light on how modern AI works\, identify potential failures\, and yield principled methods for improving efficiency\, reliability\, and performance.\nNoam earned his PhD in Computer Science at Tel Aviv University\, where he was advised by Nadav Cohen. Prior to that\, he obtained a BSc in Computer Science (summa cum laude) at The\nHebrew University of Jerusalem under the Amirim honors program. For his research\, Noam received several honors and awards\, including the Zuckerman Postdoctoral Scholarship\, the Israeli Council for Higher Education (VATAT) Postdoctoral Scholarship\, the Apple Scholars in AI/ML PhD fellowship\, the Tel Aviv University Center for AI and Data Science excellence fellowship\, and the Deutsch Prize for PhD candidates.
URL:https://cis.haifa.ac.il/event/noam-razin-princeton-university-fundamentals-of-aligning-general-purpose-ai/
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260122T110000
DTEND;TZID=Asia/Jerusalem:20260122T120000
DTSTAMP:20260416T022324
CREATED:20260105T064758Z
LAST-MODIFIED:20260105T064758Z
UID:5081-1769079600-1769083200@cis.haifa.ac.il
SUMMARY:Orr Fischer (Bar Ilan University) - A Global View of Locality Through the Lens of Distributed Computing\, Parallel Computing\, and Learning Theory
DESCRIPTION:Title: A Global View of Locality Through the Lens of Distributed Computing\, Parallel Computing\, and Learning Theory \nAbstract:\nLocality\, i.e. computing under partial knowledge\, is a fundamental challenge that can manifest itself in many versatile ways: from a distributed network where each processor has to rely on local information to solve global tasks (such as routing)\, to a sublinear algorithm that only has access to a small fraction of the input before outputting some estimate or approximate solution. In this talk\, I will highlight the many faces of this challenge in the distributed\, parallel\, and learning settings. \n– Distributed settings: CONGEST is a distributed model which aims to capture both locality and communication limitations. I will present several results for this model\, which highlight the interplay of the two limitations\, and in particular the subgraph freeness problem – a very local problem that requires a surprisingly global solution. Additionally\, I will discuss how to overcome challenges arising from the inclusion of other considerations\, such as security\, and resilience to crashes and corruptions. \n– Parallel settings: The widely adopted Map-Reduce Framework revolutionized parallel computation in the industry\, and has been the focus of much theoretical research. The Massively-parallel computation model (MPC) is one such avenue to explore this framework. I will discuss several results in the MPC model\, and in particular an interesting new research direction\, in which we show that even a single machine with a slightly more global view is sufficient to overcome many of the known lower bounds for this setting. At the heart of our techniques are sampling lemmas designed to connect small but informative parts of the input (e.g. random subgraphs if the input is a graph) to global properties of the input. \n– Learning Theory: In the PAC learning setting\, we would like to be able to answer queries about a set of data elements\, given only a few training examples randomly taken from a distribution. Through techniques developed for the study of locality and graph theory\, we are able to give new insights into various problems of interest\, such as the sample complexity of contrastive learning and hierarchical clustering\, and also obtain insights to the k nearest neighbors problem.
URL:https://cis.haifa.ac.il/event/orr-fischer-bar-ilan-university-a-global-view-of-locality-through-the-lens-of-distributed-computing-parallel-computing-and-learning-theory/
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260120T120000
DTEND;TZID=Asia/Jerusalem:20260120T130000
DTSTAMP:20260416T022324
CREATED:20260105T065749Z
LAST-MODIFIED:20260105T065749Z
UID:5087-1768910400-1768914000@cis.haifa.ac.il
SUMMARY:Mor Weiss (Bar-Ilan University)
DESCRIPTION:
URL:https://cis.haifa.ac.il/event/mor-weiss-bar-ilan-university/
LOCATION:Hanamal 65 St.\, Amir Building\, Seminar Room 413\, Hanamal 65 St.\, Amir Building\, Seminar Room 413\, Haifa\, Israel
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260115T110000
DTEND;TZID=Asia/Jerusalem:20260115T120000
DTSTAMP:20260416T022324
CREATED:20260105T064638Z
LAST-MODIFIED:20260105T064638Z
UID:5079-1768474800-1768478400@cis.haifa.ac.il
SUMMARY:Yotam Dikstein (IAS) - High Dimensional Expanders: Structure and Applications
DESCRIPTION:Title: High Dimensional Expanders: Structure and Applications \nAbstract: \nExpanders are graphs that are both edge-sparse and well connected. They have been an instrumental tool in many results in mathematics and computer science\, some of which seem to have little connection to graphs at all. High dimensional expanders (HDXs) are hypergraph analogues of expander graphs — sparse and well connected hypergraphs.  HDXs have already played a key role in several recent results in theoretical computer science and combinatorics. However\, much of their underlying theory remains unexplored. \nMotivated by this\, we will see what HDXs are\, why they generalize expander graphs\, and how they serve as a common setting for results in probability and computational complexity. \nI will illustrate their power by presenting new Chernoff-type inequalities for HDXs and explaining their applications to: \n1. Hardness amplification – constructing functions for which the best polynomial-time algorithm performs no better than random guessing. \n2. Hardness of approximation – constructing natural problems for which even approximate solutions are intractable unless P=NP. \nThe talk will not assume prior background and is aimed at a broad audience.
URL:https://cis.haifa.ac.il/event/yotam-dikstein-ias-high-dimensional-expanders-structure-and-applications/
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;VALUE=DATE:20260115
DTEND;VALUE=DATE:20260116
DTSTAMP:20260416T022324
CREATED:20251209T123818Z
LAST-MODIFIED:20260115T130844Z
UID:2958-1768435200-1768521599@cis.haifa.ac.il
SUMMARY:יום פרוייקטים סמסטר א'
DESCRIPTION:סטודנטים בשנה ג' מציגים את הפרויקט המסכם שלהם לתואר באירוע חגיגי המאורגן על ידי החוג. הצגת הפרויקטים תתקיים בבניין דילן\, קומה 2. לאחר מכן\, יעברו המשתתפים לבניין עמיר להרצאת אורח ולהכרזה על המנצחים בקטגוריות השונות של הפרויקט המצטיין. \nמצורפת הזמנה לכנס
URL:https://cis.haifa.ac.il/event/%d7%99%d7%95%d7%9d-%d7%a4%d7%a8%d7%95%d7%99%d7%99%d7%a7%d7%98%d7%99%d7%9d-%d7%a1%d7%9e%d7%a1%d7%98%d7%a8-%d7%90/
LOCATION:בניין דילן\, הנמל 16\, חיפה\, North\, Israel
CATEGORIES:אירועים מערכות מידע
ORGANIZER;CN="%D7%A9%D7%9C%D7%99 %D7%A7%D7%A8%D7%93%D7%A9%D7%98%D7%99%D7%99%D7%9F":MAILTO:skradsh1@univ.haifa.ac.il
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260108T110000
DTEND;TZID=Asia/Jerusalem:20260108T120000
DTSTAMP:20260416T022324
CREATED:20260105T064430Z
LAST-MODIFIED:20260105T064430Z
UID:5077-1767870000-1767873600@cis.haifa.ac.il
SUMMARY:Jonathan Shafer (MIT) - From Learning Theory to Cryptography: Provable Guarantees for AI
DESCRIPTION:Title: From Learning Theory to Cryptography: Provable Guarantees for AI \nAbstract: \nEnsuring that AI systems behave as intended is a central challenge in contemporary AI. This talk offers an exposition of provable mathematical guarantees for learning and security in AI systems. \nStarting with a classic learning-theoretic perspective on generalization guarantees\, we present two results quantifying the amount of training data that is provably necessary and sufficient for learning: (1) In online learning\, we show that access to unlabeled data can reduce the number of prediction mistakes quadratically\, but no more than quadratically [NeurIPS23\, NeurIPS25 Best Paper Runner-Up]. (2) In statistical learning\, we discuss how much labeled data is actually necessary for learning—resolving a long-standing gap left open by the celebrated VC theorem [COLT23]. \nProvable guarantees are especially valuable in settings that require security in the face of malicious adversaries. The main part of the talk adopts a cryptographic perspective\,  showing how to: (1) Utilize interactive proof systems to delegate data collection and AI training tasks to an untrusted party [ITCS21\, COLT23\, NeurIPS25]. (2) Leverage random self-reducibility to provably remove backdoors from AI models\, even when those backdoors are themselves provably undetectable [STOC25]. \nThe talk concludes with an exploration of future directions concerning generalization in generative models\, and AI alignment against malicious and deceptive AI.
URL:https://cis.haifa.ac.il/event/jonathan-shafer-mit-from-learning-theory-to-cryptography-provable-guarantees-for-ai/
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260106T120000
DTEND;TZID=Asia/Jerusalem:20260106T130000
DTSTAMP:20260416T022324
CREATED:20260105T065546Z
LAST-MODIFIED:20260105T065855Z
UID:5084-1767700800-1767704400@cis.haifa.ac.il
SUMMARY:Omrit Filtser (Open University) -  Robustly Guarding Polygons
DESCRIPTION:Title: Robustly Guarding Polygons \nAbstract: A fundamental set cover problem that arises in geometric domains is the classic Art Gallery Problem: given a geometric domain (e.g.\, a polygon)\, place a set of points within the domain\, such that every point in it is seen by at least one of the guards. This problem has many variants and has been studied extensively from many perspectives\, including combinatorics\, complexity\, approximation algorithms\, and algorithm engineering.\nIn this talk I will propose precise notions of what it means to robustly guard a domain\, under a variety of models. While approximation algorithms for minimizing the number of (precise) point guards in a polygon is a notoriously challenging area of investigation\, we show that imposing various degrees of robustness on the notion of visibility coverage leads to a more tractable (and realistic) problem for which we can provide approximation algorithms with constant factor guarantees. \nBased on joint work with Rathish Das\, Matya Katz\, and Joe Mitchell.
URL:https://cis.haifa.ac.il/event/omrit-filtser-open-university-robustly-guarding-polygons/
LOCATION:Hanamal 65 St.\, Amir Building\, Seminar Room 413\, Hanamal 65 St.\, Amir Building\, Seminar Room 413\, Haifa\, Israel
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Jerusalem:20260105T110000
DTEND;TZID=Asia/Jerusalem:20260105T120000
DTSTAMP:20260416T022324
CREATED:20260105T063751Z
LAST-MODIFIED:20260105T063751Z
UID:5029-1767610800-1767614400@cis.haifa.ac.il
SUMMARY:Tamer Mour (Bocconi University) - Computing on Encrypted Data; Extremely Fast and Simple
DESCRIPTION:Title: Computing on Encrypted Data; Extremely Fast and Simple \n\nAbstract: \nConsider the following setting for computing on private data. A client uploads an encryption of a big input X to an untrusted server\, and then wishes to make an unbounded number of queries f(X) while hiding f and X from the server and using only its secret key. How efficiently can this be done?\nI will present a new framework for achieving the above for two useful special cases:\n1- Data access (secret-key private information retrieval; sk-PIR)\, where X is a database and f(X)=X[i] for a secret index i.\n2- Linear computations (encrypted matrix-vector product; EMVP)\, where X is a matrix and f(X)=Xv for a secret vector v.\nOur framework yields extremely practical solutions\, often approaching the concrete efficiency of cleartext computation across relevant metrics (down to a x1.25 overhead). This is enabled by a non-traditional approach to cryptographic design that prioritizes real-world considerations and from which new sources of hardness emerge.\nIn addition\, we obtain new feasibility results for sk-PIR and EMVP based on the standard Learning Parity with Noise assumption (LPN)\, in a parameter regime not known to imply public-key encryption. \nBased on joint works with Fabrice Benhamouda\, Caicai Chen\, Yuval Ishai\, Shai Halevi\, Hugo Krawczyk\, Tal Rabin and Alon Rosen.
URL:https://cis.haifa.ac.il/event/cs-seminar-computing-on-encrypted-data/
CATEGORIES:סמינרים מדעי המחשב
END:VEVENT
END:VCALENDAR