# ChunkAct preprint reports improved robot task success with fewer LLM calls

_Published Wednesday, October 14, 2026 at 1:50 AM EDT · Science, Robotics · Latest · Tier 2 — Notable_

Researchers report that ChunkAct improved robot task success by 5.0 points over the state of the art in agentic robotics on LIBERO-Pro while requiring 2.9x fewer large language model calls. The preprint also reports 4.4x fewer input tokens and 2.3x fewer output tokens.

The method compiles a robot's memory into sequences of actions. A single model call starts a sequence, which runs autonomously and returns control upon failure. This reserves model reasoning for decisions instead of individual actions. The reported gains come from the LIBERO-Pro benchmark.

## Sources

- [arXiv Query: search_query=cat:cs.RO&id_list=&start=0&max_results=30](https://arxiv.org/abs/2610.18874v1)

---
Canonical: https://techandbusiness.org/newswire/4sgFAa-XRn1T2K0UYspu3d
Published: 2026-10-14T05:50:58.428Z
Story chronology: 2026-10-13T17:59:57.000Z
Retrieved: 2026-10-14T08:59:04.556Z
Publisher: Tech & Business (techandbusiness.org)