A Clean Answer Is Not a Clean Agent: What an Unlearning Certificate Covers Once the Model Calls Tools, Why the Agentic Leaks Divide Into Residue, Scope, Reacquisition and Exposure, and the Source-by-Observer Audit That Would Assign Each to Its Owner

APA

Mahendrakar, P. (2026). A Clean Answer Is Not a Clean Agent: What an Unlearning Certificate Covers Once the Model Calls Tools, Why the Agentic Leaks Divide Into Residue, Scope, Reacquisition and Exposure, and the Source-by-Observer Audit That Would Assign Each to Its Owner [Preprint]. Life of Research. https://doi.org/10.5281/zenodo.23766691

Copy APA

BibTeX

@misc{lor1612026,
  title        = {A Clean Answer Is Not a Clean Agent: What an Unlearning Certificate Covers Once the Model Calls Tools, Why the Agentic Leaks Divide Into Residue, Scope, Reacquisition and Exposure, and the Source-by-Observer Audit That Would Assign Each to Its Owner},
  author       = {Pranay Mahendrakar},
  year         = {2026},
  publisher    = {Life of Research},
  doi          = {10.5281/zenodo.23766691},
  url          = {https://lifeofresearch.com/preprints/161},
  keywords     = {machine unlearning, LLM agents, tool use, retrieval-augmented generation, reasoning traces, privacy leakage, evaluation, right to be forgotten},
  note         = {LOR-00161}
}

Copy BibTeX More on the record page