What a Deletion Certificate Covers, and Where It Expires: Auditable Removal from a Support-Vector Memory
Not on the map. Why?
Abstract (arXiv)
We study deletion in a context memory that fits a support-vector boundary around stored keys and uses the resulting nonnegative coefficients to weight their values. An exactly zero coefficient lets us remove a key without changing the current normalized readout. In a three-key construction, however, a later admission makes the discarded key active in a solve over the full history. For a positive-weight key, a decremental update targets a refit on the remaining keys at the original coefficient cap, under its stated assumptions. Earlier standalone comparisons show close typical gate-score agreement, while similar keys with different values can yield larger readout differences. The journal extension tests chained deletions and admissions, checking coefficient feasibility and optimality conditions in both returned states and fresh references. The initial audit completes 66 of 80 trajectories and exposes numerical failures. A separately checked wrapper completes all 5,120 scheduled operations at unchanged acceptance thresholds, with eight initialization rescues and 84 update rescues. Its largest readout difference is 0.5273% of retained-value range. These finite results support explicit acceptance and rescue rules for the evaluated memory edits. They do not extend the current-state removal guarantee to arbitrary future admissions.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 14 of its 22 references that are in the map. First posted on arXiv 13 Jul 2026.
Builds on
14 of 22 references.
References · 14
- 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,718
- 2020Transformers are RNNs: Fast Autoregressive Transformers with Linear Attentioncited by 2,022
- 2019Machine Unlearningcited by 1,061
- 2024SnapKV: LLM Knows What You are Looking for Before Generationcited by 880
- 2023H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Modelscited by 832
- 2017OptNet: Differentiable Optimization as a Layer in Neural Networkscited by 599
- 2021Linear Transformers Are Secretly Fast Weight Programmerscited by 502
- 2019Differentiable Convex Optimization Layerscited by 465
- 2023TinyStories: How Small Can Language Models Be and Still Speak Coherent English?cited by 449
- 2024Parallelizing Linear Transformers with the Delta Rule over Sequence Lengthcited by 396
- 2024Titans: Learning to Memorize at Test Timecited by 328
- 2023Uncovering mesa-optimization algorithms in Transformerscited by 100
2 more
Built on by
Citing papers per year: 2026: 2.
2 papers.