The paper shows that deterministic cache eviction cannot ensure consistent serving-time error estimation and proposes a randomized approach to restore identifiability and provide error certificates.

This paper presents SCP, a cache partitioning design that combines strict eviction isolation with write-shared coherence to mitigate eviction-based cache side channels.

Leyline introduces a novel serving-side primitive that allows agentic LLMs to perform targeted, efficient edits to the KV cache, avoiding costly full re-prefilling after content modification.

The paper demonstrates that structural protection mechanisms are the dominant factor in maintaining high performance for KV cache eviction policies, often surpassing the benefits of complex scoring al…

The paper proposes AsymCache, a computation-latency-aware KV cache management system that optimizes LLM inference by aligning cache eviction decisions with GPU attention kernel performance, significan…

This paper investigates the persistence of a timing side channel through the shared page cache in modern cloud platforms with various isolation mechanisms.

PURGE is a novel machine unlearning algorithm that leverages the duality between continual learning and unlearning to achieve high data retention while making the unlearned model indistinguishable fro…

The paper characterizes 'dead-entry' TLB misses in GPUs, which occur when recently evicted translations are immediately re-walked, and proposes DEPOT, a Bloom filter mechanism that significantly reduc…

A simple online algorithm is presented for the stacking problem to avoid shifts with a sufficient condition involving stacking area dimension, load/unload points, and maximum items.

The paper proposes GroundedCache, an evidence-validated cache router that significantly improves the safety of reusing cached semantic answers in RAG systems by requiring multiple gates to validate th…

The paper introduces MemSecBench, a benchmark for evaluating the lifecycle security of agent memory systems against malicious semantics, and reports results across various configurations.

This paper introduces CW-Ghost, a method for estimating cache line fill volume and determining helper-thread prefetching granularity based on cache capacity constraints.

FreeMOCA is a memory- and compute-efficient continual learning framework that uses adaptive layer-wise interpolation in parameter space to prevent catastrophic forgetting when analyzing evolving malwa…

The paper proposes FOAM, an adaptive damping method that stabilizes the Shampoo optimization algorithm by dynamically controlling damping and eigendecomposition frequency, thereby reducing staleness-i…

The Sleator-Tarjan model used in competitive analysis of paging is incorrect, and this error undermines the performance predictions for online paging algorithms.

MemMorph introduces a novel memory poisoning attack that biases LLM agent tool selection by injecting crafted records into the agent's long-term memory, achieving high success rates even against moder…

The paper proposes adaptive caching for key-value (KV) caching in large language models, improving cache hit rate and reducing time to first token.

The paper analyzes the bit-flip vulnerability of shared KV-cache blocks in LLM serving systems, demonstrating that these blocks are susceptible to silent, persistent, and selective data corruption.

This paper introduces a new lock-free multi-word compare-and-swap algorithm with a contention-aware helping mechanism and version embedding approach to improve efficiency and consistency, achieving up…

CachePrune introduces a privacy-aware, fine-grained KV cache sharing mechanism that allows LLM inference systems to safely reuse cache entries across users' requests, significantly improving efficienc…