NanoKnow partitions QA questions by pre-training data presence to separate the effects of memorized facts from external evidence in LLM outputs.
Data-faithful feature attribution: Mitigating unobservable confounders via instrumental variables.Advances in Neural Information Processing Systems, 37:44935–44964, 2024a
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
A single adversary in distributed training inflates its attribution value via latent optimization on synthetic batches without degrading accuracy or triggering basic defenses.
citing papers explorer
-
NanoKnow: How to Know What Your Language Model Knows
NanoKnow partitions QA questions by pre-training data presence to separate the effects of memorized facts from external evidence in LLM outputs.
-
On the Fragility of Data Attribution When Learning Is Distributed
A single adversary in distributed training inflates its attribution value via latent optimization on synthetic batches without degrading accuracy or triggering basic defenses.