mlsanityData to 14 Oct 2026

Specification-Aware Distribution Shaping for Robotics Foundation Models

arXiv 2603.17969 ↗First posted 18 Mar 2026cs.RO cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Robotics foundation models have demonstrated strong capabilities in executing natural language instructions across diverse tasks and environments. However, they remain largely data-driven and lack formal guarantees on safety and satisfaction of time-dependent specifications during deployment. In practice, robots often need to comply with operational constraints involving rich spatio-temporal requirements such as time-bounded goal visits, sequential objectives, and persistent safety conditions. In this work, we propose a specification-aware action distribution optimization framework that enforces a broad class of Signal Temporal Logic (STL) constraints during execution of a pretrained robotics foundation model without modifying its parameters. At each decision step, the method computes a minimally modified action distribution that satisfies a hard STL feasibility constraint by reasoning over the remaining horizon using forward dynamics propagation. We validate the proposed framework in simulation using a state-of-the-art robotics foundation model across multiple environments and complex specifications.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 23 of its 28 references that are in the map. First posted on arXiv 18 Mar 2026.

Builds on

23 of 28 references.

References · 23

  1. 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
  2. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  3. 2022Objaverse: A Universe of Annotated 3D Objectscited by 1,540
  4. 2017AI2-THOR: An Interactive 3D Environment for Visual AIcited by 869
  5. 2023Foundation Models in Robotics: Applications, Challenges, and the Futurecited by 216
  6. 2023Toward General-Purpose Robots via Foundation Models: A Survey and Meta-Analysiscited by 103
  7. 2025SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learningcited by 72
  8. 2023Plug in the Safety Chip: Enforcing Constraints for LLM-driven Robot Agentscited by 69
  9. 2016Q-Learning for Robust Satisfaction of Signal Temporal Logic Specificationscited by 67
  10. 2024PoliFormer: Scaling On-Policy RL with Transformers Results in Masterful Navigatorscited by 58
  11. 2024Grammar-Aligned Decodingcited by 54
  12. 2024FLaRe: Achieving Masterful and Adaptive Robot Policies with Large-Scale Reinforcement Learning Fine-Tuningcited by 54
11 more
  1. 2023SPOC: Imitating Shortest Paths in Simulation Enables Effective Navigation and Manipulation in the Real Worldcited by 48
  2. 2025Generating Robot Constitutions & Benchmarks for Semantic Safetycited by 33
  3. 2020Tractable Reinforcement Learning of Signal Temporal Logic Objectivescited by 22
  4. 2015Robust Temporal Logic Model Predictive Controlcited by 21
  5. 2024SELP: Generating Safe and Efficient Task Plans for Robot Agents with Large Language Modelscited by 21
  6. 2014On-Line Monitoring for Temporal Logic Robustnesscited by 15
  7. 2022Deep reinforcement learning under signal temporal logic constraints using Lagrangian relaxationcited by 8
  8. 2022Funnel-based Reward Shaping for Signal Temporal Logic Tasks in Reinforcement Learningcited by 8
  9. 2023Multi-Agent Reinforcement Learning Guided by Signal Temporal Logic Specificationscited by 7
  10. 2025Constrained Decoding for Safe Robot Navigation Foundation Modelscited by 5
  11. 2022Control Barrier Functions with Actuation Constraints under Signal Temporal Logic Specificationscited by 2

Built on by

Not cited yet.