LLM Digest

Story

arxiv_cs_lg · Jun 18, 2026 · paper

Source brief

Execution-State Capsules: Graph-Bound Execution-State Checkpoint and Restore for Low-Latency, Small-Batch, On-Device Physical-AI Serving

arxiv.orgJun 18, 2026
original source linked

In brief

Mainstream LLM serving systems reuse prefix work mainly through paged or radix key-value (KV) caches. This is highly effective for high-throughput, high-concurrency serving, but it manages only one positional fragment...

Feed lens

agenteval

Read the original at arxiv.org →Open in live feed Read that day’s brief

Execution-State Capsules: Graph-Bound Execution-State Checkpoint and Restore for Low-Latency, Small-Batch, On-Device Physical-AI Serving

Earlier in this thread 4 items

Decoupling Inference from State Updates in Low-Latency Feature Engines via Probabilistic Thinning

Gemma 4 12B Enables On-Device, Multimodal Agentic Workflows with an Encoder-free Architecture

Vortex: Efficient and Programmable Sparse Attention Serving for AI Agents

Stateful Inference for Low-Latency Multi-Agent Tool Calling