RAVEN: Receiver-Conditioned Action-Value Encoding for Finite-Alphabet Multi-Agent Communication
Abstract
A message drawn from a small alphabet helps a teammate only if it keeps the distinctions that change that teammate's next decision. We show that scoring messages by action values averaged over the receiver's situation can erase exactly these distinctions, and we propose RAVEN (Receiver-conditioned Action-Value ENcoding), which trains a four-symbol, one-step-delayed channel to preserve each receiver's centered action-value profile within the receiver's own context. The sender never needs to know that context: the receiver decodes every symbol with its private information. We give two estimators of this target. With a teacher, offline RAVEN selects the codebook that exactly minimizes an empirical conditional distortion and distills it into a frozen sender; we bound the resulting codebook-selection error and one-step decision loss. Without a teacher, online RAVEN aligns, inside a QMIX learner, the deployed symbol pathway with a training-only continuous reference that shares its routing. Against five recent communication methods on eight navigation settings, offline RAVEN attains the highest return in seven, and removing receiver conditioning forfeits 83% of its communication gain. Online RAVEN raises predator-prey capture success from 53.2% to 96.0% over the same QMIX backbone without communication, and on SMAC and MPE it attains the best mean normalized score of 14 methods, including methods that exchange kilobit messages. Every RAVEN message costs 2 bits, 12-1,024x fewer than those of NDQ, CACOM and ExpoComm on navigation.
Related
- Class-Conditioned Diffusion Augmentation and Explainable In-Context Ensemble Learning for Imbalanced Maintenance Action Classification in Aircraft Systems
- The Expected Number of Distinct Substrings in an Alphabet String
- Maximizing Measures for Subadditive and Almost-Additive Potentials on Countable-Alphabet Subshifts
- Alphabet-Dependent Bounds for Pure Quantum (r,ρ)-Locally Recoverable Codes
- Robust Reconstruction on Trees with a Growing Alphabet
BibTeX
@misc{sun2026raven,
title = {{RAVEN: Receiver-Conditioned Action-Value Encoding for Finite-Alphabet Multi-Agent Communication}},
author = {Shuwei Sun and Chenxi Wang and Jian Huang and Weiyun Ru and Hui Cao},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.37566},
eprint = {2609.37566},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.37566},
}
RIS
TY - UNPB
TI - RAVEN: Receiver-Conditioned Action-Value Encoding for Finite-Alphabet Multi-Agent Communication
AU - Shuwei Sun
AU - Chenxi Wang
AU - Jian Huang
AU - Weiyun Ru
AU - Hui Cao
PY - 2026
DA - 2026/09/29
PB - arXiv
DO - 10.48550/arxiv.2609.37566
UR - https://arxiv.org/abs/2609.37566
AB - A message drawn from a small alphabet helps a teammate only if it keeps the distinctions that change that teammate's next decision. We show that scoring messages by action values averaged over the receiver's situation can erase exactly these distinctions, and we propose RAVEN (Receiver-conditioned Action-Value ENcoding), which trains a four-symbol, one-step-delayed channel to preserve each receiver's centered action-value profile within the receiver's own context. The sender never needs to know that context: the receiver decodes every symbol with its private information. We give two estimators of this target. With a teacher, offline RAVEN selects the codebook that exactly minimizes an empirical conditional distortion and distills it into a frozen sender; we bound the resulting codebook-selection error and one-step decision loss. Without a teacher, online RAVEN aligns, inside a QMIX learner, the deployed symbol pathway with a training-only continuous reference that shares its routing. Against five recent communication methods on eight navigation settings, offline RAVEN attains the highest return in seven, and removing receiver conditioning forfeits 83% of its communication gain. Online RAVEN raises predator-prey capture success from 53.2% to 96.0% over the same QMIX backbone without communication, and on SMAC and MPE it attains the best mean normalized score of 14 methods, including methods that exchange kilobit messages. Every RAVEN message costs 2 bits, 12-1,024x fewer than those of NDQ, CACOM and ExpoComm on navigation.
ER -