PhD student @WisconsinCS, MS @CMU_ECE. Advised by @jhasomesh. Work on Trustworthy Machine Learning
Madison, WI
Joined October 2017
- Tweets31
- Following184
- Followers89
- Likes80
Jihye Choi retweeted
A @NobelPrize for a Badger 🏅
Congratulations to Francis Halzen, named the 2026 Nobel laureate in physics for his groundbreaking work with @uw_icecube to advance our understanding of neutrinos and the universe’s most energetic events.
news.wisc.edu/university-of-…
Jihye Choi retweeted
Have you ever wondered: In federated learning, what if we could leverage clients' private data without compromising privacy—what more could we achieve?
🚀 We're excited to introduce SLVR (Securely Leveraging Client Validation for Robust Federated Learning).
Check out the full paper for details: arxiv.org/pdf/2502.08055
Why SLVR? Building on secure Multi-party Computation (MPC), SLVR offers a fresh perspective on combining privacy and robustness in federated learning:
Jihye Choi retweeted
Very happy to hear that GANs are getting the test of time award at NeurIPS 2024.
The NeurIPS test of time awards are given to papers which have stood the test of the time for a decade.
I took some time to reminisce how GANs came about and how AI has evolve in the last decade.
Jihye Choi retweeted
Differential Privacy and RAG in @llama_index 🔐
There are so many datasets that would be important to share and learn from with LLMs/RAG systems, but unfortunately contain sensitive individual information:
🏥 Healthcare: Doctors and researchers can study patient data to find patterns/treatments - but there’s individual patient details
🌐 Online sites: Companies collect data to improve their offerings - but individuals may want to keep their identity private
@_nerdai_ has led the charge to build the first privacy-preserving RAG network implementation. As a data provider, synthetically generate copies of your data without leaking individual information, and offer it up to a RAG service for others to query.
Check out our full blog post here - llamaindex.ai/blog/retrievin…
Prviacy-Safe-Retrieval demo: github.com/run-llama/llama_i…
Differential privacy llamapack: llamahub.ai/l/llama-packs/ll…
Notebooks for the differential privacy llamapack: github.com/run-llama/llama_i…
🚀 Excited to present our work at MLHC 2024 tomorrow in beautiful Toronto 🇨🇦
Join us as we introduce MALADE—the first effective explainable multi-agent LLM system for extracting Adverse Drug Events (ADEs) from FDA drug labels and drug prescription data.
5/n
If you're curious about how we've achieved this, I’d love to connect at the session and dive into the details!
🗓 Poster Session: August 16th, 11:15 AM - 12:15 PM @mlforhc
📄 Paper: arxiv.org/abs/2408.01869
💻 Code: lnkd.in/gwNyMYPm
1
146
n/n
A huge thank you to my incredible collaborators—Nils Palumbo (co-lead), @pchalasani, Matthew Engelhard, @jhasomesh , Anivarya Kumar, and David Page—from @WisconsinCS , @DukeMedSchool, and Langroid ✨
154
Jihye Choi retweeted
Our work on Pharmacovigilance using multi-agent LLM powered system. This was joint work with Duke medical and especially proud of the students for pulling this off. Blog post by @pchalasani about the paper.
Paper is going to appear in MLHC 2024.
Paper: arxiv.org/abs/2408.01869
Blog:
langroid.github.io/langroid/…
Jihye Choi retweeted
I feel like a lot of people leverage LLMs suboptimally, especially for long-form interactions that span a whole project. So I wrote a VSCode extension that supports what I think is a better use paradigm. 🧵 1/N
Extension: marketplace.visualstudio.com…
Code: github.com/locuslab/chatllm-…
Jihye Choi retweeted
Looking to recruit PhD students (in both ECE + CS @ UIUC) with STRONG interests in ML+Sec, Applied Crypto, Systems, and demystifying foundation models. Deadline: Dec 15. Ping with any questions. Please RT and help amplify?
Jihye Choi retweeted
I'm on the academic job market
I'm a full-stack data privacy researcher
I build systems that are 1)provably private 2)functionally-rich 3)compatible w/ real-world constraints
I do this by exploring the synergy between cryptography & differential privacy, both in theory & practice
Jihye Choi retweeted
We are pleased to publish a paper based on the findings of the workshop held on June 27 and organized by Google, Stanford, and UW-Madison.
arxiv.org/abs/2308.14840
We hope that this paper will be useful for the community and thus welcome your feedback.
Jihye Choi retweeted
@CadeMetz at the New York Times just published a piece on a new paper we are releasing today, on adversarial attacks against LLMs. You can read the piece here:
nytimes.com/2023/07/27/techn…
And find more info and the paper at: llm-attacks.org/ [1/n]
Jihye Choi retweeted
Check out our ICML 2023 paper “Stratified Adversarial Robustness with Rejection”, where we analyze adversarially-robust classification with a reject option.
Paper: arxiv.org/abs/2305.01139
Code: github.com/jfc43/stratified-…
Can we interpret why your out-of-distribution (OOD) detector thinks an input is in-distribution or OOD in terms of human-understandable concepts? Then check out our new @ICML2023 paper, "Concept-based Explanations for Out-Of-Distribution Detectors"
arxiv.org/pdf/2203.02586.pdf
I deeply thank my amazing collaborators, Jayaram Raghuram, Ryan Feng, @jiefengchen1, @jhasomesh and @atulprakash. @WisconsinCS @UMichCSE
See you at Honolulu!
258
5/n
👩💻 Code: github.com/jihyechoi77/conce…
🌴 Poster session: July 26, 11 am - 12:30 pm HST icml.cc/virtual/2023/poster/…
1
202