Jack Merullo
Research Scientist, Goodfire · 1 appearance on the record.
computed by AI from the episodes · how this works → · full disclaimer →
scientistacademicjmerullo.github.io ↗
Jack Merullo is a research scientist at Goodfire focusing on mechanistic interpretability and model representations. He earned his Ph.D. in Computer Science from Brown University, where he studied representation learning in deep neural networks.
1 opinion · 1 insight · every statement was checked. none of them is a claim the record can settle: opinions, insights and disclosures never carry an assessment.
Everything Jack Merullo said on Latent Space that made the record, most notable first. Filter by type, assessment or year in the ledger →
Merullo: Current machine unlearning techniques merely suppress data rather than removing it
“I would describe it more as not unlearning, but maybe suppression. I think there's, like, really, like, I guess, guarantees that you've fully removed information from a model is, is, I don't think it's been convincingly showed anywhere yet”
Merullo: LLM memorization spans a gradient from reasoning to rote recall
“You can actually see, like the way that we, like, disentangle memorization, you can kind of see this like, gradient of memorization in between both mechanistically and behaviorally with, like, logical reasoning tasks being quite distinct from rote memorization…”
Appearances (1)
| Episode | Date | Speaking time |
|---|---|---|
| [State of MechInterp] SAEs in Production, Circuit Tracing, AI4Science, "Pragmatic" Interp | Dec 31, 2025 | 7m |