Kernel-Managed Shared Memory for System-Wide Personalization
AI systems become more useful when they can adapt to the people using them, but in multi-agent systems, useful context learned by one agent often remains unavailable to others. We present kernel-managed shared memory, a system-level abstraction in which specialized agents write structured, tagged memories while the agent-system kernel, not individual agents, governs retrieval, privacy enforcement, and prompt injection. We implement and evaluate this design on AIOS and compare it against three alternatives across three assistant models (GPT-4o, Llama-3.1:8B, Qwen-2.5:7B) and 1,800 total trials. Against an unmanaged external memory backend (Mem0) using identical underlying storage, kernel-managed retrieval and injection improve personalization scores by 2.4-4.0 points on a 5-point scale (e.g., 1.05 to 4.69 profile usage on GPT-4o), with every comparison significant at p < 10^-18. Against standard retrieval-augmented injection, gains are similarly large and consistent across all three models. Against full, unfiltered context concatenation, a soft ceiling on available context rather than on response quality, kernel-managed injection statistically matches performance on two of three models and shows a small, model-specific deficit on the third, while using substantially shorter prompts: end-to-end latency is 15-61% lower across all three models, with corresponding reductions in per-call token usage and inference cost. These results indicate that centralizing memory management in the agent-system kernel, rather than leaving retrieval and privacy enforcement to individual agents, delivers most of the personalization benefit of unconstrained context at a fraction of its cost.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
PAT: Accelerating LLM Decoding via Prefix-Aware Attention with Resource Efficient Multi-Tile Kernel
LLM serving is increasingly dominated by decode attention, which is a memory-bound operation due to massive KV cache loading from global memory. Meanwhile, real-world workloads exhibit substantial, hierarchical shared pr…
Fast NF4 Dequantization Kernels for Large Language Model Inference
Large language models (LLMs) have grown beyond the memory capacity of single GPU devices, necessitating quantization techniques for practical deployment. While NF4 (4-bit NormalFloat) quantization enables 4$\times$ memor…
QUICK: Quantization-aware Interleaving and Conflict-free Kernel for efficient LLM inference
We introduce QUICK, a group of novel optimized CUDA kernels for the efficient inference of quantized Large Language Models (LLMs). QUICK addresses the shared memory bank-conflict problem of state-of-the-art mixed precisi…
GPUQuantizationThe 2021 Urdu Fake News Detection Task using Supervised Machine Learning and Feature Combinations
This paper presents the system description submitted at the FIRE Shared Task: "The 2021 Fake News Detection in the Urdu Language". This challenge aims at automatically identifying Fake news written in Urdu. Our submitted…
Fake News DetectionLemmatizationResource-aware Computation-Communication Overlap for multi-GPU ML Workloads
The rapid growth of large-scale machine learning (ML) has made distributed training across multiple GPUs a fundamental component of modern ML systems. As model sizes and computational throughput continue to increase, com…