Research
I am a postdoc in the Data Science for Humanity Group at the Max Planck Institute for Security and Privacy in Germany, working with Meeyoung Cha. I study how large language models encode knowledge in their parameters and activations, and how they decode it during inference.
Memory, unlearning, and editing
Localizing the parameters that hold a specific memory, and testing whether it can be truly erased or changed without damaging unrelated abilities, and whether the change generalizes beyond the targeted memory. (AI Engram, ICML 2026 Oral; Erase or Hide?, ICLR 2026; Moir, NeurIPS 2026)
Out-of-context reasoning
How models infer facts their training data implies but never states, without verbalizing their thoughts. For example, why models trained on “A is B” often fail to infer “B is A”, and which internal representations fix this. (Bilinear representation mitigates reversal curse, ICLR 2026)
Previously, I did my PhD in physics at KAIST, advised by Prof. Hawoong Jeong at the Complex Systems and Statistical Physics Lab, using deep learning to study nonequilibrium statistical physics.
News
-
Moir, on preserving reasoning under knowledge editing by estimating the preservation covariance from the model’s own samples, was accepted at NeurIPS 2026.
-
Presented AI research using agents and notebooks as one of five speakers at Spark(l)ing Science, MPI-SP, in the session AI Agents, Show and Tell: Real Workflows, Real Speedups.
-
AI Engram, on identifying memory traces in neural networks, was selected for an Oral at ICML 2026 (168 orals, top 0.7% of submissions).
-
Recognized as a Gold Reviewer for ICML 2026, placing among the conference’s top reviewers.
Selected publications
-
AI Engram: In Search of Memory Traces in Artificial Intelligence
-
Transformer as a hippocampal memory consolidation model based on NMDAR-inspired nonlinearity
NeurIPS 2023Paper
Contact
dong-kyum.kim [at] mpi-sp.org