JHU Computer Science

Talking robots learn to manage human interruptions Johns Hopkins computer scientists designed an interruption-handling system to facilitate more natural conversations with social robots

Johns Hopkins researchers including @victor-antony.bsky.social and @ziangxiao.bsky.social‬ have created a system that could make social robots more effective at detecting and managing user interruptions—a breakthrough for areas like health care and education where natural conversation is crucial.

In the world of AI, who speaks for the trees? Johns Hopkins computer scientists’ new training method, EigenLoRAx, promises to cut the environmental cost of fine-tuning large AI models.

Johns Hopkins computer scientists have developed a method that dramatically cuts the environmental costs of fine-tuning #AI. They call it EigenLoRAx and, like the Dr. Seuss character, it’s a nod to a greener future. 🌳

“I gotta be honest, the whole thing is pretty nuts,” says JHU CS’ @matthewdgreen.bsky.social. “My best hypothesis is the phones are seeing a Wi-Fi access point that is slightly off from where they are.”

Websites are tracking you via browser fingerprinting New research provides first evidence of the use of browser fingerprints for online tracking.

Clearing your cookies is not enough to protect your privacy online, say researchers from @jhu.edu and @tamu.bsky.social.

Joanne Selinski promoted to teaching professor. Computer Science.

Congratulations to Joanne Selinski on her promotion to teaching professor! 🥳 Also our associate head for education & director of undergraduate studies, Prof. Selinski’s research interests lie in computer science education, particularly as it relates to core courses & active learning approaches. 🎓

Finding the virtue in AI Feeling uneasy about the rise of AI? Microsoft data scientist and JHU CS alum Juan M. Lavista Ferres finds virtue in the technology and is more bullish about than bullied by its rise.

Johns Hopkins Magazine spoke to CS alum Juan M. Lavista Ferres about working for @microsoft.com and the potential for #AI applications to solve some of the world’s most pressing issues. Read the Q&A:

New faculty Q&A: Michael Oberst Get to know Michael Oberst, who joins Johns Hopkins as an assistant professor of computer science and a member of the Malone Center for Engineering in Healthcare.

Learn more about @moberst.bsky.social’s research, courses, and hobbies in our new Q&A!

Fluent, Fast, and Fair - Johns Hopkins Whiting School of Engineering In collaboration with Microsoft Research, Johns Hopkins computer scientists have built a new machine translation model that achieves top-tier performance across 50 diverse languages.

Fluent, Fast, Fair: Recent @jhucompsci.bsky.social PhD grad Haoran Xu & team built a new machine translation model that achieves top performance across 50 diverse languages WITHOUT needing bloated models. The trick is "adaptive rejection preference optimization." engineering.jhu.edu/news/fluent-...

Finding the virtue in AI Juan M. Lavista Ferres, the chief data scientist at Microsoft's AI for Good Lab, discusses how artificial intelligence might solve the world's most pressing issues

Feeling uneasy about the rise of #AI? @microsoft.com data scientist and JHU CS alum Juan M. Lavista Ferres finds virtue in the technology and is more bullish about than bullied by its rise.

Will Iran retaliate? US officials warn of cyber attack risks after strikes Federal officials and cybersecurity experts have warned of a heightened risk of cyber attacks.

“Iran is not as much of a threat as the guys that are trying to get money out of people’s bank accounts,” says JHU CS’ Anton Dahbura. “That's what people should be worried about, realistically.”

Headshots of Steve Rifkin, Steve DeBlasio, Chris Venghaus, and Piper Gold.

Happy #SysAdminDay to our amazing IT team! There’s no better day than this to familiarize yourself with their awesome wiki: support.cs.jhu.edu/wiki/Main_Page 😉

It’s AI vs. the ATS Johns Hopkins computer science students are fighting back against a job hunter’s worst enemy: the applicant tracking system.

Inspired by their own experiences with the online recruitment process, six Johns Hopkins CS students are fighting automated resume scanners with an AI assistant of their own: Jobiter, an intelligent job search platform that helps find the best job opportunities tailored to users’ specific profiles.

Democratizing Creativity - Johns Hopkins Whiting School of Engineering In their quest to develop a co-creative AI partner, Johns Hopkins computer scientists discover valuable insights about the limitations and ethical considerations of current generative AI models.

Will generative AI push out creators... or empower them? ID.8—from @jhucompsci.bsky.social PhD student Victor Nikhil Antony—is designed as a co-creative agent, emphasizing human control and input to enhance storytelling, not hijack it. engineering.jhu.edu/news/democra...

Don’t scan that scam! QR code fraud is rising, so a Johns Hopkins team built a system to sniff out fakes.

Johns Hopkins researchers have created an AI-powered weapon against one of the internet’s fastest-growing threats: QR code scams that have skyrocketed by more than 2,400% over the last five years.

Ben Langmead promoted to full professor. Computer science.

Congratulations to @benlangmead.bsky.social on his promotion to full professor! 🎉 Prof. Langmead is recognized across the computational and life sciences fields for his innovative methods helping to transform how biomedical researchers and other life scientists access and use DNA sequencing data. 🧬

Scientists Identify Genes For Tomato And Eggplant Size Geneticists mapping nightshade genomes used CRISPR gene editing to grow bigger fruits without sacrificing flavor.

Researchers mapping the genomes of nightshades used CRISPR gene editing to grow bigger tomatoes and eggplants without sacrificing flavor. Dr. @mikeschatz.bsky.social joins us to discuss this research and the current state of genetically modified crops.

Gregory D. Hager to serve as interim director for LCSR He assumes the role from Russell Taylor, who stepped down after leading the Laboratory for Computational Sensing and Robotics for over a decade.

Thanking Russ for his years of service and guidance and welcoming Greg into this new leadership role!

Congrats! Kim Franklin 2025 WSE Staff Excellence Award.

Kim Franklin and other staff members pose with their awards next to Dean Ed Schlesinger.

Kim Franklin and other staff members pose with Dean Ed Schlesinger.

Congratulations to our PhD academic program coordinator Kim Franklin on winning a @hopkinsengineer.bsky.social Staff Excellence Award! This is a well-deserved recognition of her invaluable contributions to the dept. & the enthusiasm, leadership, & unwavering support she provides to our PhD students!

Speak and your X-ray will be imaged Researchers from the Johns Hopkins University’s ARCADE Lab present the voice-controlled X-ray imaging system that earned a Best Paper Award at IPCAI 2025.

Researchers in @jhu.edu’s ARCADE Lab—including @bdkilleen.bsky.social & @mathias-unberath.bsky.social—have developed an #IPCAI2025 Best Paper Award-winning, voice-controlled #AI system that can image and analyze patient X-rays in real time.

+ @henrikvonkleist.bsky.social, Ilya Shpitser, @www.helmholtz-munich.de’s @carstenmarr.bsky.social, & researchers from @ruhr-uni-bochum.de will be presenting “Feature Importance Metrics in the Presence of Missing Data,” which introduces the distinction between 2 eval. frameworks under missing data:

JHU CS’ Yinzhi Cao recommends using virtual card numbers and using a separate browser for financial transactions. Learn more about protecting your financial info online: money.usnews.com/credit-cards...

CS Faculty Research Panel. Labeled headshots of Michael Dinitz, Kimia Ghobadi, and Jessica Sorrell.

📢 Attention undergrads! 📢 Check out our 2nd CS Faculty Research Panel of the summer on Thursday, July 31—faculty like @mdinitz.bsky.social & @optimistsinc.bsky.social will help you learn how to apply your degree to the broader world of CS—at JHU & beyond. Learn more at www.cs.jhu.edu/event/comput...

The TriageNow and RescueReady teams holding their awards and posing with Larry Aronhime at HopStart 2025.

Subhasri Vijay, Jamie Stelnik, and Larry Aronhime pose together.

The TriageNow team: Kamila Wong, Meghana Karthic, Maria-Noelia Herne, and Micaela Rodriguez Steube.

Last semester, computer science teams tackled critical gaps in #EMS training and mass casualty response with award-winning, innovative digital platforms. Learn more: www.cs.jhu.edu/news/from-cl...

Yinzhi Cao receives multiple awards at IEEE Security and Privacy 2025 He received both a Distinguished Paper Award and a Test of Time Award at the premier forum for presenting developments in computer security and electronic privacy.

Congratulations to Yinzhi Cao on the awards at #IEEESP25!

Congratulations! Ramchandran Muthukumar successfully defended his dissertation “From Parsimony to Local Sensitivity: Generalization and Robustness in Modern Machine Learning” under the guidance of advisor Jeremias Sulam. Ramchandran plans to join the Data Science Institute at the University of Chicago as a postdoctoral scholar. We in the department are extremely proud of our students who have successfully completed their PhD. Congratulations on this achievement and best wishes as you begin an exciting new phase of life!

Congratulations, @ram-cha.bsky.social!

Welcome to JHU CS! 💙

WATCH: Adaptive Monitoring for AI Deployments via Weighted-Conformal Martingales Responsibly deploying artificial intelligence (AI) / machine learning (ML) systems in high-stakes settings arguably requires not only proof of system reliability, but also continual, post-deployment m...

& “WATCH: Adaptive Monitoring for AI Deployments via Weighted-Conformal Martingales” by @drewprinster.bsky.social, @aliu33.bsky.social, and more proposes a weighted generalization of conformal test martingales to lay a theoretical foundation for online monitoring for unexpected changepoints: (5/5)

Raman Arora and Thanh Nguyen-Tang present a learning algorithm that helps a decision-maker perform well even when facing an opponent that learns and adapts over time in their poster “Policy Regret Minimization in Markov Games with Function Approximation”: (4/5)

Yunjuan Wang and Raman Arora’s poster “Backdoor Attacks in Token Selection of Attention Mechanism” examines how backdoor attacks exploit transformer models’ attention mechanisms: (3/5)

Intersectional Fairness in Reinforcement Learning with Large State and Constraint Spaces In traditional reinforcement learning (RL), the learner aims to solve a single objective optimization problem: find the policy that maximizes expected reward. However, in many real-world settings, it ...

‪@optimistsinc.bsky.social and researchers from @pennengineering.bsky.social present “Intersectional Fairness in Reinforcement Learning with Large State and Constraint Spaces,” which generalizes the problem of maximizing the reward of a minimum reward group: (2/5)