- photo - C.V. - google scholar - a piece of art I made

I am a Ph.D. candidate in computer science at Boston University, under the supervision of professor Ran Canetti. I'll defend in June 2024. August 2024, I am joining Harvard University as a postdoc in the SAFR AI lab led by professor Seth Neel and professor Salil Vadhan.

I am interested in theoretical computer science and machine learning (and especially their intersection). My Ph.D. thesis discovers new relationships between the complexity theory of circuit lower bounds and computational learning theory. Most recently, I've sought to explain phenomena like the success of multimodal data in AI/ML. I've also done theoretical study on undetectable model stealing attacks.

Increasingly, I'm interested in more applied research in AI/ML. Things like data attribution in generative modelling, machine unlearning, and AI/ML safety in particular. I'll be working on these areas at Harvard.

















Publications

(ab) indicates author names listed alphabetically.

  • Karchmer, Ari. On Stronger Computational Separations Between Multimodal and Unimodal Machine Learning. ICML 2024. ArXiv preprint.
  • Karchmer, Ari. Agnostic Membership Query Learning with Nontrivial Savings: New Results and Techniques. ALT 2024. ArXiv preprint.
  • Karchmer, Ari. Distributional PAC-Learning from Nisan's Natural Proofs. ITCS 2024. Winner of best student paper award. Invited for publication at TheoretiCS. ArXiv preprint.
  • Karchmer, Ari. Theoretical Limits of Provable Security Against Model Extraction by Efficient Observational Defenses. SaTML 2023. IACR ePrint.
  • Canetti, Ran and Karchmer, Ari. Covert Learning: How to Learn with an Untrusted Intermediary. TCC 2021. Invited to Journal of Cryptology. IACR ePrint. (ab)

Teaching fellowships


Select talks

  • "Undetectable Model Stealing with Covert Learning" - Harvard University SAFR AI Lab, Cambridge MA, April '24 Slides
  • "Cryptography and Complexity Theory in the Design and Analysis of ML" - Vector Institue, Toronto CA, April '24 Slides
  • "Learning from Nisan's Natural Proofs" - MIT CIS Seminar, March '24 Slides
  • "Distributional PAC-learning from Nisan's Natural Proofs" - ITCS, Simons Institute, Jan '24 Video
  • "Undetectable Model Stealing and more with Covert Learning" - Google Research MTV, algorithms seminar, Jan '24 Slides
  • "New Approaches to Heuristic PAC-Learning vs. PRFs" - Lower Bounds, Learning, and Average-case Complexity Workshop at Simons Institute, Feb '23 Simons talk
  • "The Limits of Provable Security Against Model Extraction" - Privacy Preserving Machine Learning Workshop at Crypto, Aug '22 PPML talk
  • "Covert Learning: How to Learn with an Untrusted Intermediary" - Charles River Crypto Day at MIT, Nov '21 Crypto day talk



Flag Counter