# Efficient Mixture-of-Agents Serving via Tree-Structured Routing, Adaptive Pruning, and Dependency-Aware Prefill-Decode Overlap

arXiv: https://arxiv.org/abs/2512.18126 · HTML: https://mlsanity.com/paper/2512.18126/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 31 references that are in the map.
- First posted on arXiv 19 Dec 2025.

## Builds on (most cited first, 18 of 18)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [Tree of Thoughts: Deliberate Problem Solving with Large Language Models](https://mlsanity.com/paper/2305.10601/) (2023)
- [Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate](https://mlsanity.com/paper/2305.19118/) (2023)
- [Graph of Thoughts: Solving Elaborate Problems with Large Language Models](https://mlsanity.com/paper/2308.09687/) (2023)
- [DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving](https://mlsanity.com/paper/2401.09670/) (2024)
- [ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMs](https://mlsanity.com/paper/2309.13007/) (2023)
- [SARATHI: Efficient LLM Inference by Piggybacking Decodes with Chunked Prefills](https://mlsanity.com/paper/2308.16369/) (2023)
- [Cut the Crap: An Economical Communication Pipeline for LLM-based Multi-Agent Systems](https://mlsanity.com/paper/2410.02506/) (2024)
- [Examining Inter-Consistency of Large Language Models Collaboration: An In-depth Analysis via Debate](https://mlsanity.com/paper/2305.11595/) (2023)
- [Agentic Reward Modeling: Integrating Human Preferences with Verifiable Correctness Signals for Reliable Reward Systems](https://mlsanity.com/paper/2502.19328/) (2025)
- [A Comprehensive Analysis of Large Language Model Outputs: Similarity, Diversity, and Bias](https://mlsanity.com/paper/2505.09056/) (2025)
- [Rethinking Semantic Parsing for Large Language Models: Enhancing LLM Performance with Semantic Hints](https://mlsanity.com/paper/2409.14469/) (2024)
- [CORE: Measuring Multi-Agent LLM Interaction Quality under Game-Theoretic Pressures](https://mlsanity.com/paper/2508.11915/) (2025)
- [Mining Useful General Data for Low-Resource Domain Adaptation](https://mlsanity.com/paper/2511.07380/) (2025)
- [Assessing the similarity of real matrices with arbitrary shape](https://mlsanity.com/paper/2403.17687/) (2024)

## Abstract (arXiv)

Mixture-of-Agents (MoA) inference can suffer from dense inter-agent communication and low hardware utilization, which jointly inflate serving latency. We present a serving design that targets these bottlenecks through an algorithm-system co-design. First, we replace dense agent interaction graphs with a hierarchical tree topology that induces structured sparsity in inter-agent communication. Second, we introduce a runtime adaptive mechanism that selectively terminates or skips downstream agent invocations using semantic agreement and confidence signals from intermediate outputs. Third, we pipeline agent execution by overlapping incremental prefilling with decoding across dependency-related agents, improving utilization and reducing inference latency. Across representative tasks, this approach substantially reduces end-to-end latency (up to 90%) while maintaining comparable accuracy (within ±1%) relative to dense-connectivity MoA baselines, and can improve accuracy in certain settings.