The paper reports fine‑grained, deterministic instability in Dynamic Tensor Rematerialization (DTR), an online eviction policy for memory‑constrained DNN training. On an LSTM trace, tiny changes in memory budget (0.10% of peak) switch the system between fast and slow execution regimes with up to 7.3× overhead differences, driven by repeated re‑eviction of the same storages. On a ResNet‑32 trace, a deterministic feasibility inversion is observed: the run is feasible at a 0.101 budget ratio, infeasible (OOM) between 0.102–0.106, and feasible again from 0.107, caused by a fully pinned recursive rematerialization frontier exceeding the budget after all evictable tensors are removed. The authors attribute the LSTM instability to the joint size‑staleness scoring term and argue that these represent two distinct budget‑sensitive pathologies rather than a single mechanism.
By Mahesh Reddy Pagadala
arXiv:2607. 07144v1 Announce Type: new Abstract: The key-value (KV) cache dominates the memory cost of long-context autoregressive inference, and a growing body of work compresses it through quantization, eviction, or offloading.
By Vladimir Gusev
arXiv:2607. 12204v2 Announce Type: replace Abstract: Auditable memory requires a precise contract: which output is preserved, relative to which reference solve, and across which updates.
By Vishwajith Ramesh
arXiv:2607. 03701v1 Announce Type: cross Abstract: Large language models (LLMs) can propose circuit-optimization decisions, but industrial analog flows cannot expose foundry PDK content, proprietary schematics, absolute simulation paths, or license-bound tool state to a cloud endpoint.
By Xunqi Li, Chris H. Kim
arXiv:2609.27981v1 Announce Type: cross
Abstract: KV-cache eviction is typically evaluated through average quality-memory trade-offs, yet a small average loss can hide requests whose utility degrades...
By Beomgu Kang, SoJin Yun, Hojoon Kim, Hyunseok Seo
arXiv:2608. 05863v1 Announce Type: new Abstract: Modern models no longer keep a plain KV cache: latent caches, learned sparse selectors and recurrent states each carry the model's memory in a different form, and each fails differently under compression.
By Fanzhe Wei, Li Liu, Ziyang Wang, Chenyu Wang
arXiv:2607. 19623v1 Announce Type: cross Abstract: We characterize per-bit-position fault sensitivity in ML inference across 16 workloads -- spanning transformer-based models and attention-free CNNs -- and across three floating-point formats.
By Muhammad Husnain Mubarik, Karthik Mohan Kumar, Pedro Antonio Pena, Keshavan Varadarajan, Kunal Tyagi
arXiv:2610.19498v1 Announce Type: cross
Abstract: Model signing authenticates the full-precision weights W an LLM is released with, yet users run a quantized artifact Q whose producing transformation...
By Wei-Liang Liu, Cheng-Yen Tsai
arXiv:2610.03027v1 Announce Type: cross
Abstract: The key-value (KV) cache becomes a major memory bottleneck in long-context LLM inference, placing substantial pressure on memory capacity and bandwid...
By Minsoo Cheong, Donghyun Son, Sungjoo Yoo
arXiv:2608. 20280v1 Announce Type: cross Abstract: Semantic caches reuse an LLM response when the incoming query embedding lies near a cached query, but proposed eviction policies have rarely been compared under one protocol.
By Yash Kulkarni, Shubham Harkare, Arvind Suresh Yogesh Babu
arXiv:2609.37887v1 Announce Type: new
Abstract: Activation and key-value cache precision change what a quantized language model computes without altering its stored weights. Direct weight-code bounds...
By Arian Eamaz, Mojtaba Soltanalian
arXiv:2606. 20128v1 Announce Type: cross Abstract: Benchmarks for LLM-generated GPU kernels (KernelBench, TritonBench, GEAK) score correctness through fixed-shape, small-sample allclose-style checks.
By Dipankar Sarkar