Trendoday

Evolving Infra: Agent-Adaptive LLM Serving at Scale

Agentic workloads are a major source of LLM inference demand. Existing prefill--decode (PD) disaggregated serving systems offer limited flexibility for runtime adaptation of compute allocation and KV

  • Upvotes12
–Rank now
#13Peak rank
5Days on list
285Score · 7d
Nov 12, 2026First seen
Rank · 7 days

About this paper

Agentic workloads are a major source of LLM inference demand. Existing prefill--decode (PD) disaggregated serving systems offer limited flexibility for runtime adaptation of compute allocation and KV cache management, and rely on coarse signals that only partly capture the diverse and dynamic agent workload. These limitations lead to load imbalance between prefill and decode nodes and eviction of KV caches that agents will reuse. We present Evolving Infra, an agent-adaptive serving framework that adapts compute allocation to agent workloads and KV cache management to agent context management at runtime. We decouple compute allocation from key--value (KV) cache management, allowing nodes to…

Upvotes
12
Authors
11
Published
Nov 9, 2026

Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.

Trending history

Evolving Infra: Agent-Adaptive LLM Serving at Scale was first seen on Hugging Face · Daily papers on Nov 12, 2026. Since then it has appeared in 44 checks on 5 different days. Its best position was #13 of 30, reached on Nov 13, 07:28 UTC. It dropped off the list after Nov 16, 03:28 UTC.

Rank history

Position on Hugging Face · Daily papers at every check (lower is better).

#1#11#20#30 Nov 12Nov 13Nov 14Nov 15Nov 16 Nov 12, 13:28 UTC · #23Nov 12, 15:28 UTC · #22Nov 12, 17:28 UTC · #20Nov 12, 19:28 UTC · #18Nov 12, 21:28 UTC · #17Nov 12, 23:28 UTC · #18Nov 13, 01:28 UTC · #17Nov 13, 03:28 UTC · #17Nov 13, 05:28 UTC · #17Nov 13, 07:28 UTC · #13Nov 13, 09:28 UTC · #17Nov 13, 11:28 UTC · #17Nov 13, 13:28 UTC · #18Nov 13, 15:28 UTC · #18Nov 13, 17:28 UTC · #20Nov 13, 19:28 UTC · #20Nov 13, 21:28 UTC · #19Nov 13, 23:28 UTC · #18Nov 14, 01:28 UTC · #18Nov 14, 03:28 UTC · #19Nov 14, 05:28 UTC · #19Nov 14, 07:28 UTC · #20Nov 14, 09:28 UTC · #20Nov 14, 11:28 UTC · #20Nov 14, 13:28 UTC · #20Nov 14, 15:28 UTC · #20Nov 14, 17:28 UTC · #20Nov 14, 19:28 UTC · #20Nov 14, 21:28 UTC · #20Nov 14, 23:28 UTC · #20Nov 15, 01:28 UTC · #20Nov 15, 03:28 UTC · #20Nov 15, 05:28 UTC · #20Nov 15, 07:28 UTC · #20Nov 15, 09:28 UTC · #20Nov 15, 11:28 UTC · #20Nov 15, 13:28 UTC · #20Nov 15, 15:28 UTC · #20Nov 15, 17:28 UTC · #20Nov 15, 19:28 UTC · #20Nov 15, 21:28 UTC · #20Nov 15, 23:28 UTC · #21Nov 16, 01:28 UTC · #21Nov 16, 03:28 UTC · #18

Last 14 days

Best rank per day (UTC).

Nov 3today

Metric history

Platform metric as reported by Hugging Face.

04913 Nov 12Nov 13Nov 14Nov 15Nov 16 Nov 12, 13:28 UTC · 6Nov 12, 15:28 UTC · 7Nov 12, 17:28 UTC · 9Nov 12, 19:28 UTC · 11Nov 12, 21:28 UTC · 11Nov 12, 23:28 UTC · 11Nov 13, 01:28 UTC · 11Nov 13, 03:28 UTC · 11Nov 13, 05:28 UTC · 12Nov 13, 07:28 UTC · 13Nov 13, 09:28 UTC · 13Nov 13, 11:28 UTC · 13Nov 13, 13:28 UTC · 13Nov 13, 15:28 UTC · 13Nov 13, 17:28 UTC · 12Nov 13, 19:28 UTC · 12Nov 13, 21:28 UTC · 12Nov 13, 23:28 UTC · 12Nov 14, 01:28 UTC · 12Nov 14, 03:28 UTC · 12Nov 14, 05:28 UTC · 12Nov 14, 07:28 UTC · 12Nov 14, 09:28 UTC · 12Nov 14, 11:28 UTC · 12Nov 14, 13:28 UTC · 12Nov 14, 15:28 UTC · 12Nov 14, 17:28 UTC · 12Nov 14, 19:28 UTC · 12Nov 14, 21:28 UTC · 12Nov 14, 23:28 UTC · 12Nov 15, 01:28 UTC · 12Nov 15, 03:28 UTC · 12Nov 15, 05:28 UTC · 12Nov 15, 07:28 UTC · 12Nov 15, 09:28 UTC · 12Nov 15, 11:28 UTC · 12Nov 15, 13:28 UTC · 12Nov 15, 15:28 UTC · 12Nov 15, 17:28 UTC · 12Nov 15, 19:28 UTC · 12Nov 15, 21:28 UTC · 12Nov 15, 23:28 UTC · 12Nov 16, 01:28 UTC · 12Nov 16, 03:28 UTC · 12

Questions

Is Evolving Infra: Agent-Adaptive LLM Serving at Scale trending on HF Daily Papers right now?

Not at the moment. It was last on the list on Nov 16, 03:28 UTC, after 5 days of trending.

What is the highest rank Evolving Infra: Agent-Adaptive LLM Serving at Scale reached?

#13 on Hugging Face · Daily papers, on Nov 13, 07:28 UTC. Trendoday checks this list every 120 minutes.

How long has Evolving Infra: Agent-Adaptive LLM Serving at Scale been trending?

It was first seen on Nov 12, 2026 and has been on the list on 5 days (44 checks) since.