ArcticSwarm: Deferring Early Consensus in Long-Horizon Multi-Agent Research

Soyoung Yoon, Boyi Liu, Yite Wang, Ruofan Wu, Canwen Xu, Nikki Lijing Kuang, Seung-won Hwang, Yuxiong He, Zhewei YaoarXiv 2026 · arXiv

TL;DR

ArcticSwarm is a novel multi-agent architecture designed for long-horizon web research that prevents "premature consensus"—the tendency of AI agents to copy each other's early, incorrect guesses. By isolating agents during evidence gathering and enforcing rigorous, multi-stage reviews before committing to an answer, it significantly outperforms existing deep-research baselines, achieving up to 88.3% accuracy on complex research benchmarks.

What Problem Does This Paper Address?

Multi-agent systems work beautifully when there is an automated way to check their work, such as running unit tests on generated code. But what happens when you ask a swarm of AI agents to conduct open-ended web research where no simple "verifier" exists?

Typically, developers rely on majority voting or allow agents to chat freely to reach a consensus. However, the authors identify a critical failure mode in this setup: premature consensus. When agents share partial findings too early, they anchor on the same initial (and often incorrect) hypothesis. This collapses the diversity of the search space; agents stop exploring alternative angles and collectively fail along the same path.

Methods

The authors introduce ArcticSwarm, an architecture that decouples evidence gathering from evidence integration using two main pillars:

  1. Gated Bulletin Board System (BBS): Instead of free-flowing peer-to-peer communication, all agent interactions are routed through a centralized BBS. Crucially, the orchestrator can dynamically assign agents to Isolation Mode (where they can write findings to the board but cannot read peer posts) or Collaboration Mode (where they can read and write). This ensures agents explore independent search paths without being biased by early group-think.
  2. Staged Review at Commitment Boundaries: Before any hypothesis is accepted, it must pass through three strict gates: an internal Self-Check, a Board Audit by a dedicated auditor agent, and a final Commit Gate. The commit gate requires a "VERIFIED" verdict from both the original investigating agent and an independent reviewer, alongside an explicit sweep for alternative candidates.
Figure 2: ArcticSwarm architecture. The orchestrator posts tasks to a gated Bulletin Board System (BBS) and spawns subagents with different profiles and ac- cess modes. Agents assigned to isolation mode cannot read the BBS, whereas agents assigned to collaboration mode can read and write. Example traces in Figure 7.

Datasets

The architecture was evaluated on two challenging long-horizon web search benchmarks:

  • BrowseComp: A dataset containing 1,266 complex, multi-constraint identification questions evaluated against the live web.
  • BrowseComp-Plus: An upgraded version containing 830 questions evaluated against a static, pre-defined corpus of ~100K curated web documents to ensure perfect evaluation replicability.

Key Insights and Ideas

  • Communication Can Be Harmful: In open-ended research, dense communication actually hurts performance by accelerating convergence and reducing search diversity. Restricting what agents can read during active search forces them to find unique evidence.
  • The Answer is Usually Found, Just Not Chosen: A failure analysis revealed that the correct answer is actually present somewhere in the agents' collective browser history in 58% of failed cases. The breakdown happens during the adjudication phase when choosing between candidates.
  • Forcing Contrarian Thinking Works: Forcing the orchestrator to launch at least one "alternative/contrarian" search task dramatically boosts accuracy, breaking the candidate fixation loop.

Key Results

  • State-of-the-Art Accuracy: ArcticSwarm sets a new benchmark on BrowseComp-Plus, achieving 82.6% accuracy using the open-weight Qwen 3.5-27B model, beating the strong MiroFlow baseline (70.6%).
  • Frontier Model Performance: When scaled to GPT-5 on the live-web BrowseComp, ArcticSwarm hits 73.6%, vastly outperforming OpenAI's own deep-research baseline (54.9%) and MiroFlow (63.4%).
  • Ablation Proof: Turning off gated isolation dropped accuracy by 3.8 percentage points, and disabling the review gates dropped it by another 4.3 percentage points, proving both components are vital.
Figure 1: ArcticSwarm sets the strongest BrowseComp- Plus accuracy across three model families (GPT-5, Son- net 4.5, and open-weight Qwen 3.5-27B), beating both each provider’s deep-research system and the strongest open swarm (MiroFlow) under matched configurations (Table 2).

Limitations and Future Work

  • High Token Cost: The extensive multi-stage review process and independent search paths consume significantly more tokens than simpler architectures. Optimizing token and resource efficiency is left for future work.
  • Live-Web Flakiness: Performance on the live web is heavily bottlenecked by upstream infrastructure issues. Gateway authentication errors, rate limits, and safety filters accounted for roughly 91% of subagent task failures.
  • Heuristic Orchestration: The policy for when to switch an agent between isolation and collaboration mode is currently heuristic and could be made more dynamic or learned in future iterations.

Summary updated

Abstract

Multi-agent systems have shown strong performance in domains with reliable verifiers such as coding, where multi-parallel candidate generation selected by a verifier is effective. However, such pipelines would not generalize to open-ended, long-horizon research tasks without a verifier. While majority voting or self-consistency is often used to reach consensus as a proxy verifier, parallel agents repeatedly explore the same evidence, while access to peers' partial findings cause search to converge on an early candidate before alternatives are tested. We present ArcticSwarm, a multi-agent research architecture that separates evidence gathering from evidence integration. Subagents publish findings to a shared bulletin board, while gated isolation lets selected search tasks maintain their own prior, preventing early consensus. Structured review at three commitment boundaries enforce only confident candidates to be propagated. As a result, ArcticSwarm reaches 82.6% on the full BrowseComp-Plus set with the open-weight Qwen 3.5-27B model, compared with 78.8% without gated isolation and 74.5% additionally with structured review disabled, outperforming aligned baseline MiroFlow runs (70.6%). Extending to live-web BrowseComp, ArcticSwarm reaches 73.6% with GPT-5, which is well above the reported provider system (54.9%) and MiroFlow (63.4%). Overall, the results show that restricting peer reads during evidence gathering and strengthening commitment boundaries before a hypothesis is shared can broaden search and improve long-horizon multi-agent deep research.
BibTeX
@misc{yoon2026arcticswarm,
  author = {Soyoung Yoon and Boyi Liu and Yite Wang and Ruofan Wu and Canwen Xu and Nikki Lijing Kuang and Seung-won Hwang and Yuxiong He and Zhewei Yao},
  title = {{ArcticSwarm: Deferring Early Consensus in Long-Horizon Multi-Agent Research}},
  year = {2026},
  eprint = {2609.01870},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.01870},
}
Figure