preprint open access

Four Things Called Forgetting: Interference, Transience, Reset and Unlearning Remove Different Things, Why Forgetting Looks Easy by Accident and Hard on Purpose Only Under Different Instruments, and the Savings Measurement That Would Tell Them Apart

Abstract

Life of Research

(c) 2026 Pranay Mahendrakar. Licensed under CC BY 4.0.

Machine learning uses one word for four operations. Catastrophic forgetting is damage that fine-tuning does to earlier capabilities. Transience is the fading of individual training examples during ordinary training. Resets deliberately reinitialise part of a network to restore its ability to learn. Unlearning deliberately removes targeted knowledge. Read side by side, the literatures appear to contradict each other: forgetting is too easy to cause by accident, as when ten fine-tuning examples strip a model's safety behaviour, and too hard to cause on purpose, as when unlearned knowledge returns after a few steps of fine-tuning on unrelated data. This paper argues that the contradiction comes largely from the instruments. Accidental forgetting is usually scored by output accuracy, and deliberate forgetting by adversarial recovery. When accidental forgetting is probed the same way, most of it is recoverable too: in one controlled study task accuracy falls from near 100 percent to about 20 percent while a recovery probe still reaches 96 percent. The paper separates three things a forgetting operation can change: access (whether stored content reaches the output), content (whether it survives cheap recovery) and trainability (how fast the network learns anything new). Gradient-based forgetting, accidental or deliberate, mostly changes access. The operations shown so far to change content are mainly ones that do not carry the trained parameters forward: retraining without the data, distilling into a fresh or noised network, or filtering the data before pretraining. Plasticity resets use the same lever. The author consolidates and analyses the published measurements, states a decision procedure, and proposes a two-ratio savings protocol that would place any forgetting result on the partition.

Pranay Mahendrakar, AI specialist

About the author

Pranay Mahendrakar is an ai specialist and large language model engineer based in Bengaluru, India. He builds production artificial intelligence systems and publishes open-access research on how those systems fail. See all 161 papers by Pranay Mahendrakar, or his ORCID record.

Questions about this paper

Who wrote "Four Things Called Forgetting"?

Pranay Mahendrakar wrote "Four Things Called Forgetting: Interference, Transience, Reset and Unlearning Remove Different Things, Why Forgetting Looks Easy by Accident and Hard on Purpose Only Under Different Instruments, and the Savings Measurement That Would Tell Them Apart", published 28 Sep 2026. Pranay Mahendrakar is an Indian AI specialist and LLM engineer based in Bengaluru, India. He is the Managing Director of SonyTech, Nodal Coordinator at IIRS-ISRO, and an instructor at Tutorials Point. His work covers large language models, natural language processing, computer vision and retrieval-augmented generation. He publishes open-access research papers and is the author of three books: Just AI With Pranay, Multiverse of AI and It's Me LLM.

Is "Four Things Called Forgetting" free to read?

Yes. "Four Things Called Forgetting" by Pranay Mahendrakar is open access under a Creative Commons Attribution 4.0 licence, with the full PDF available from Zenodo at https://doi.org/10.5281/zenodo.23004075. There is no paywall and no account required.

How do I cite "Four Things Called Forgetting"?

Cite the DOI: Mahendrakar, P. (2026). Four Things Called Forgetting: Interference, Transience, Reset and Unlearning Remove Different Things, Why Forgetting Looks Easy by Accident and Hard on Purpose Only Under Different Instruments, and the Savings Measurement That Would Tell Them Apart. Zenodo. https://doi.org/10.5281/zenodo.23004075 A BibTeX entry is provided on this page.

Related research by Pranay Mahendrakar

← All papers by Pranay Mahendrakar