Integrated Calendar

Over the last five years my lab has explored the hypothesis that people suffering from major depressive disorder show pathological decision-making. In a series of experiments we demonstrate that the psychological “reference point” against which all hedonic experience is benchmarked is represented in the Anterior Cingulate Cortex of the Monkey. In parallel work, Helen Mayberg’s group has shown that the severity of a patient’s depression can be decoded from activity in this same area. We used this information and foraging theory to develop a behavioral tool for measuring the reference point in humans and found that a 3 minute version of our task can be used to diagnose depression with the same accuracy as a 60m clinical interview. The implications of this finding for our understanding of the mechanism of depression will be discussed.

Text-based 2D image editing models have recently reached an impressive level of maturity, motivating a growing body of work that uses them to drive 3D edits. While effective for appearance-based modifications, such 2D-centric 3D editing pipelines often struggle with fine-grained 3D editing, where localized structural changes must be applied while strictly preserving an object’s overall identity.

To address this limitation, we propose Prox-E, a training-free framework that enables fine-grained 3D control through an explicit, primitive-based geometric abstraction. Our framework first abstracts an input 3D shape into a compact set of geometric primitives. A pretrained vision-language model then edits this abstraction to specify primitive-level changes, which are subsequently used to guide a 3D generative model. This enables fine-grained, localized modifications while preserving unchanged regions of the original shape.

Through extensive experiments, we show that Prox-E consistently balances identity preservation, shape quality, and instruction fidelity more effectively than existing approaches, including 2D-based 3D editors and training-based methods.

Bio:

Etai Sella is a fourth-year PhD student at Tel Aviv University, supervised by Hadar Averbuch-Elor and Or Patashnik. His research focuses on making generative AI more controllable and editable, with an emphasis on 3D editing. He is currently an intern at Snap Research.

The input to the Multiway Cut problem is a weighted undirected graph, with nonnegative edge weights, and $k$ designated terminals. The goal is to partition the vertices of the graph into~$k$ parts, each containing exactly one of the terminals, such that the sum of weights of the edges connecting vertices in different parts of the partition is minimized. The problem is APX-hard for $k\ge3$. The currently best-known approximation algorithm for the problem for arbitrary~$k$, obtained by Sharma and Vondr\'ak [STOC 2014] more than a decade ago, has an approximation ratio of 1.2965. We present an algorithm with an improved approximation ratio of 1.2787. Also, for small values of $k \ge 4$ we obtain the first improvements in 25 years over the currently best approximation ratios obtained by Karger, Klein, Stein, Thorup, and Young [STOC 1999]. (For $k=3$ an optimal approximation algorithm is known.)

Our main technical contributions are new insights on rounding the LP relaxation of C{\u{a}}linescu, Karloff, and Rabani [STOC 1998], whose integrality ratio matches Multiway Cut's approximability ratio, assuming the Unique Games Conjecture [Manokaran, Naor, Raghavendra, and Schwartz, STOC 2008]. First, we introduce a generalized form of a rounding scheme suggested by Kleinberg and Tardos [FOCS 1999] and use it to replace the Exponential Clocks rounding scheme used by Buchbinder, Naor, and Schwartz [STOC 2013] and by Sharma and Vondr\'ak. Second, while previous algorithms use a mixture of two, three, or four basic rounding schemes, each from a different family of rounding schemes, our algorithm uses a computationally-discovered mixture of hundreds of basic rounding schemes, each parametrized by a random variable with a distinct probability distribution, including in particular many different rounding schemes from the same family. We give a completely rigorous analysis of our improved algorithms using a combination of analytical techniques and interval arithmetic.

Joint work with Joshua Brakensiek, Neng Huang and Aaron Potechin.

The plant microbiome plays a vital role in host fitness. However, the complexity of plant systems makes it difficult to disentangle the roles of individual bacterial species and their interactions with the host. Here, we developed two screening approaches using Chlamydomonas reinhardtii as a model for investigating bacterial pathogenicity and host immunity. First, we measured the effect of ~120 bacterial strains previously isolated from healthy Arabidopsis thaliana roots in a halo assay. We found nine bacterial strains with inhibitory effect on C. reinhardtii growth, all of which previously demonstrated pathogenicity towards A. thaliana, which suggests conserved mechanisms. Focusing on a green lineage specific pathogenic Burkholderia strain (MF6), we revealed it exerts its antagonistic effect through a contact-dependent secretion system. We, next, employed forward genetics in both Chlamydomonas and MF6 to address the genetic basis of pathogenicity and immunity, further characterized by RNA-seq, proteomics and functional assays.In another strategy, we characterized the genetic basis of immunity in a natural habitat. We inoculated the pooled deletion mutant library in Chlamydomonas in soil samples containing various microbial communities and quantified mutant unique barcodes abundance as a proxy for mutant fitness. A promising subset of genes was identified and provides new insights into the defense strategies and potential symbiotic mechanisms employed by green algae with conservation throughout the green lineage.Altogether, these findings highlight conserved plant/alga–bacteria interactions and establish Chlamydomonas as a fascinating system for unraveling the molecular mechanisms of host–pathogen interactions across the plant superkingdom.

As networks become more programmable, they are increasingly built around flexible software components. While this programmability enables new functionality and faster innovation, it also makes network behavior harder to reason about. In this talk, I will present a research agenda that brings ideas from formal methods to programmable networks. In particular, I will present techniques that leverage programmable-network semantics for concurrency safety, traffic monitoring, and failure recovery. More broadly, this work illustrates how semantic foundations can help bring stronger correctness guarantees to modern networked systems.

Bio
Guy Amir is a Postdoctoral Researcher at Cornell University, conducting research at the intersection of formal methods, networking, and systems. He earned his Ph.D. in 2024 from the Hebrew University of Jerusalem, where he studied AI safety, focusing on formally verifying reactive AI systems and interpreting neural networks. He holds an M.Sc. in Computer Science and a B.Sc. in Computational Biology and Computer Science, both from the Hebrew University. He has received Rothschild, Fulbright, AI-Net, and Charles Clore fellowships, as well as an ICML Spotlight and KLA Award.