Knowledge Packs deliver knowledge via pre-computed KV caches with exact equivalence under causal masking, achieving zero divergences on tested questions and enabling value-based steering without training.
Agent memory below the prompt: Persistent q4 kv cache.arXiv preprint arXiv:2603.04428
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
years
2026 5representative citing papers
QKVShare enables efficient quantized KV-cache handoff for on-device multi-agent LLMs, cutting TTFT versus re-prefill across tested contexts while adaptive quantization stays competitive with uniform baselines on GSM8K.
A single shared asymmetrically compressed KV cache pool enables up to 15 concurrent LLM agents with 2.91x compression, 97.7% memory reduction, and only +0.57% perplexity increase on Llama-3-8B.
A net-value-per-byte curator governs memory lifecycle in on-device LLM agents, cutting memory 2.7x and uplink 2.4x while driving injection success to zero on task-drift benchmarks and Jetson hardware.
A taxonomy that places 18 latent-communication methods for LLM agents on WHAT/WHICH/HOW axes and extracts design trade-offs.
citing papers explorer
-
Knowledge Packs: Zero-Token Knowledge Delivery via KV Cache Injection
Knowledge Packs deliver knowledge via pre-computed KV caches with exact equivalence under causal masking, achieving zero divergences on tested questions and enabling value-based steering without training.
-
QKVShare: Quantized KV-Cache Handoff for Multi-Agent On-Device LLMs
QKVShare enables efficient quantized KV-cache handoff for on-device multi-agent LLMs, cutting TTFT versus re-prefill across tested contexts while adaptive quantization stays competitive with uniform baselines on GSM8K.
-
PolyKV: A Shared Asymmetrically-Compressed KV Cache Pool for Multi-Agent LLM Inference
A single shared asymmetrically compressed KV cache pool enables up to 15 concurrent LLM agents with 2.91x compression, 97.7% memory reduction, and only +0.57% perplexity increase on Llama-3-8B.
-
Forget to Improve: On-Device LLM-Agent Continual Learning via Budget-Curated Memory
A net-value-per-byte curator governs memory lifecycle in on-device LLM agents, cutting memory 2.7x and uplink 2.4x while driving injection success to zero on task-drift benchmarks and Jetson hardware.
-
Beyond tokens: a unified framework for latent communication in LLM-based multi-agent systems
A taxonomy that places 18 latent-communication methods for LLM agents on WHAT/WHICH/HOW axes and extracts design trade-offs.