用智能聚合器并行处理长任务，提升研究类AI的准确率
Agentic Aggregation for Parallel Scaling of Long-Horizon Agentic Tasks
- 设计聚合代理，按需检索和合成多条推理轨迹信息
- 在六项基准上平均提升5.3%，深度研究任务最高提升10.3%
- 仅需一次代理推理开销，适合高效率长程任务场景
我们研究长时序智能体任务（如智能搜索、深度研究）的并行测试时扩展方法，即并行生成多个推理轨迹并聚合为最终答案。尽管此类扩展在链式思维推理中已证明有效，但智能体任务面临独特挑战：轨迹长、多轮交互、工具增强，且输出开放。仅聚合最终答案会丢失轨迹中的丰富信息，而拼接全部轨迹又超出模型上下文窗口。为此，我们提出AggAgent，一个将并行轨迹视为环境的聚合智能体，配备轻量级工具以检查候选解并跨轨迹搜索，实现按需导航与信息融合。在六个基准和三个模型族（GLM-4.7、Qwen3.5、MiniMax-M2.5）上，AggAgent优于所有现有聚合方法——平均绝对提升5.3%，两项深度研究任务最高达10.3%——同时开销极低，聚合成本始终不超过单次代理推理。研究证实，智能体聚合是并行测试时扩展的有效且高效路径。
原文摘要 · Abstract (English)
We study parallel test-time scaling for long-horizon agentic tasks such as agentic search and deep research, where multiple rollouts are generated in parallel and aggregated into a final response. While such scaling has proven effective for chain-of-thought reasoning, agentic tasks pose unique challenges: trajectories are long, multi-turn, and tool-augmented, and outputs are often open-ended. Aggregating only final answers discards rich information from trajectories, while concatenating all trajectories exceeds the model's context window. To address this, we propose AggAgent, an aggregation agent that treats parallel trajectories as an environment. We equip it with lightweight tools to inspect candidate solutions and search across trajectories, enabling it to navigate and synthesize information on demand. Across six benchmarks and three model families (GLM-4.7, Qwen3.5, MiniMax-M2.5), AggAgent outperforms all existing aggregation methods-by up to 5.3% absolute on average and 10.3% on two deep research tasks-while adding minimal overhead, as the aggregation cost remains bounded by a single agentic rollout. Our findings establish agentic aggregation as an effective and cost-efficient approach to parallel test-time scaling.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。