# VACS: Value-Aligned Compositional Shielding for Multi-Agent Reasoning

arXiv: https://arxiv.org/abs/2609.26135 · HTML: https://mlsanity.com/paper/2609.26135/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 24 references that are in the map.
- First posted on arXiv 7 Aug 2026.

## Builds on (most cited first, 9 of 9)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Improving Factuality and Reasoning in Language Models through Multiagent Debate](https://mlsanity.com/paper/2305.14325/) (2023)
- [Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks](https://mlsanity.com/paper/1702.01135/) (2017)
- [Maximum Entropy Deep Inverse Reinforcement Learning](https://mlsanity.com/paper/1507.04888/) (2015)
- [Compositional Shielding and Reinforcement Learning for Multi-Agent Systems](https://mlsanity.com/paper/2410.10460/) (2024)
- [Co-Evolutionary Defence of Active Directory Attack Graphs via GNN-Approximated Dynamic Programming](https://mlsanity.com/paper/2505.11710/) (2025)
- [Explainable Autonomous Cyber Defense using Adversarial Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2604.04442/) (2026)
- [AgenticVM: Agentic AI for Adaptive Software Vulnerability Management](https://mlsanity.com/paper/2605.01739/) (2026)

## Abstract (arXiv)

Multi-agent reasoning systems in high-stakes domains must be both accurate and safe, yet agents often follow heterogeneous value priorities (e.g., rigor, conciseness, safety), causing conflicting recommendations. Existing methods do not jointly provide: (i) principled inference of each agent's implicit values from behavior, (ii) compositional formal safety guarantees without full online communication, and (iii) value-aware conflict resolution with faithful explanations. We present VACS (Value-Aligned Compositional Shielding), a four-layer framework addressing all three. Layer 1 learns value-dimension rewards from pairwise preferences using Bradley-Terry modeling and infers per-agent value weights via deep MaxEnt IRL. Layer 2 encodes value constraints in a Lean-inspired DSL and synthesizes compositional assume-guarantee shields for runtime safety. Layer 3 resolves disagreement through nucleolus-based credit allocation and Hamiltonian consensus optimization under long-term value constraints. Layer 4 extracts a critical reasoning path from co-state sensitivities and generates formally grounded natural-language explanations. Our contribution is primarily a unified systems design with formalized interfaces and operational guarantees at the verifier-constrained decision level, rather than a complete end-to-end formal proof of all language-model internals. In controlled proof-of-concept evaluations with role-conditioned agent panels on NEJM-AI QA, MathInstruct-Subset, and a cybersecurity incident-response benchmark (CyberSec-Eval), VACS outperforms strong baselines in accuracy (85.4%, 95.0%, and 90.0%) while reducing logical inconsistency rates to near zero.