Paper Detail

SOMA: Strategic Orchestration and Memory-Augmented System for Vision-Language-Action Model Robustness via In-Context Adaptation

Zhuoran Li, Zhiyang Li, Kaijun Zhou, Jinyu Gu

arxiv Score 14.6

Published 2026-03-25 · First seen 2026-03-27

Research Track A · General AI

Abstract

Despite the promise of Vision-Language-Action (VLA) models as generalist robotic controllers, their robustness against perceptual noise and environmental variations in out-of-distribution (OOD) tasks remains fundamentally limited by the absence of long-term memory, causal failure attribution, and dynamic intervention capability. To address this, we propose SOMA, a Strategic Orchestration and Memory-Augmented System that upgrades frozen VLA policies for robust in-context adaptation without parameter fine-tuning. Specifically, SOMA operates through an online pipeline of contrastive Dual-Memory Retrieval-Augmented Generation (RAG), an Attribution-Driven Large-Language-Model (LLM) Orchestrator, and extensible Model Context Protocol (MCP) interventions, while an offline Memory Consolidation module continuously distills the execution traces into reliable priors. Experimental evaluations across three backbone models (pi0, pi0.5, and SmolVLA) on LIBERO-PRO and our proposed LIBERO-SOMA benchmarks demonstrate that SOMA achieves an average absolute success rate gain of 56.6%. This includes a significant absolute improvement of 89.1% in long-horizon task chaining. Project page and source code are available at: https://github.com/LZY-1021/SOMA.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{li2026soma,
  title = {SOMA: Strategic Orchestration and Memory-Augmented System for Vision-Language-Action Model Robustness via In-Context Adaptation},
  author = {Zhuoran Li and Zhiyang Li and Kaijun Zhou and Jinyu Gu},
  year = {2026},
  abstract = {Despite the promise of Vision-Language-Action (VLA) models as generalist robotic controllers, their robustness against perceptual noise and environmental variations in out-of-distribution (OOD) tasks remains fundamentally limited by the absence of long-term memory, causal failure attribution, and dynamic intervention capability. To address this, we propose SOMA, a Strategic Orchestration and Memory-Augmented System that upgrades frozen VLA policies for robust in-context adaptation without parame},
  url = {https://arxiv.org/abs/2603.24060},
  keywords = {cs.RO},
  eprint = {2603.24060},
  archiveprefix = {arXiv},
}

Metadata

{}