当 Agent 能互相传递笔记时，工作效率更高

大型语言模型可以通过将问题拆分为多个子问题并行求解，而非按顺序逐一处理，来更快完成任务。然而，在将子结果汇总到单一节点的系统中，这种加速效果存在上限。研究人员设计了一种方法来消除这一限制。

最新进展： 卡内基梅隆大学由 Xuecheng Liu 和 Daman Arora 领衔的团队提出了一种名为 Message Passing Language Models（MPLMs） 的智能体执行框架。在 MPLM 中，LLM 将子问题分配给不同线程——每个线程运行各自的一份 LLM 副本——而这些线程可以彼此通信。该方法在两类结构化谜题上的求解速度均快于替代方案。

核心洞见： 在早期的并行化方法中，LLM 会将任务分解为子任务；协调器线程则创建独立线程、向其分配子任务，并收集其输出。这种安排的问题在于，协调器可能因推理、工具调用等工作而陷入繁忙状态，导致子任务不得不等待。但对于线程之间关系能够预先确定的问题，并不需要中央协调器。相反，存在关联的线程可以直接彼此通信。让线程直接通信既消除了协调器的必要性，也限制了任一单独线程所承担的总负载。

工作方式： 在 MPLM 中，模型及其迭代过程可以写入命令，以启动线程、向特定线程发送结果、等待回复或停止线程。作者构建了相应程序，用这些命令来解决谜题，并在其思考可能解法时生成文本轨迹。研究团队使用这些轨迹训练了 Qwen3-0.6B-Base。谜题包括 3-SAT 示例（判断一个布尔公式是否可以求值为真）和数独（在方形网格中填入数字；每行、每列和每个宫内使用从 1 到该行、列或宫单元格数量的数字，且任何数字不得重复）。以下描述适用于数独求解；3-SAT 的求解过程有所不同。

• 一个父线程负责跟踪谜题中已经解决的部分。
• 对于数独网格中的每个单元格，对应线程通过排除法确定正确数字（因为数独网格预先填有部分数字，且某个单元格不能重复其所在行、列或宫中已有的数字）。在线程尚未确定数字时，它会等待接收负责其所在行、列和宫内其他单元格的线程发来的数字。随着这些数字到达，线程持续排除候选数字，直到只剩一个。一旦确定数字，它便将该数字发送给父线程及相关线程，然后停止。
• 当所有线程都停止后，父线程报告最终解答。

结果： 与两种早期方法相比，MPLM 解决这些谜题的速度更快，并且每个线程消耗的 token 更少。这两种方法分别是：单线程方法，以及一种运行并行线程、但通过协调器路由其结果的智能体执行框架。

• 在求解从 4×4 到 25×25 的数独网格时，MPLM 的平均速度显著更快。例如，面对 9×9 网格，MPLM 约在 15 秒内解决了 100% 的题目，而并行方法约在 60 秒内解决了 93% 的题目。此外，随着网格规模扩大，MPLM 的每线程 token 数增长更慢。MPLM 解决了 72% 的 25×25 数独题，而另外两种方法在学会解决这些问题之前，就已触及作者设定的上下文或计算资源限制。
• 在求解包含 8 至 20 个变量的 3-SAT 公式时，MPLM 的准确率与并行方法大致相当，约为 92% 对 91%。它的平均速度略快于并行方法，但在一些样例上快得多——有时快 2.5 倍——因为模型一旦在某个线程中找到解，就可以停止其他线程。单线程方法未能运行到超过 12 个变量的问题，因为其已填满模型的上下文窗口。

是的，但是： MPLM 的效率依赖于预先知道哪些线程必须与哪些线程通信。它在通信模式固定且易于推导的场景中表现突出，例如数独和 3-SAT。作者指出，对于开放式问题，确定这一模式可能需要精心设计提示词或额外训练。此外，数独示例仅限于能够通过排除法解决的题目，即所谓的 naked singles；因此，模型从不需要进行猜测与验证，也无需在线程之间展开更广泛的推理。

重要性： 在单线程中执行任务的模型，可能在找到解答之前就填满上下文窗口。MPLM 使模型能够将工作分配给多个线程，并由这些线程共同完成任务。

我们的看法： 作者还提示两个更大规模的模型 Qwen3-30B-A3B 和 Qwen3.6-35B-A3B 使用其方法，在 LongBench-v2 长上下文推理基准上处理问题。两个模型在 MPLM 框架中都表现出更高的准确率和更快的响应速度，平均延迟约降低 2 倍。这表明，该方法或许能够泛化到相对简单的数独和 3-SAT 示例之外的推理挑战，尽管其效果似乎在较小模型上更为显著。

Large language models can work faster by dividing problems into sub-problems to be solved in parallel instead of in sequence. However, the speed-up is capped in systems that aggregate the sub-results at a single point. Researchers devised a way to remove that limit.

What’s new: A team at Carnegie Mellon University led by Xuecheng Liu and Daman Arora proposed an agentic harness called Message Passing Language Models (MPLMs). Under MPLM, an LLM distributes sub-problems among separate threads — each running its own copy of the LLM — that can communicate with one another. This approach solved two types of structured puzzles faster than alternatives.

Key insight: In earlier parallelization methods, an LLM breaks down tasks into sub-tasks while a coordinator thread spawns separate threads, assigns sub-tasks to them, and collects their output. The problem with this arrangement is that the coordinator can become bogged down in reasoning, tool calls, and the like, so the subtasks must wait. But problems in which the relationships between threads are known in advance don’t require a central coordinator. Instead, related threads can communicate with one another. Enabling threads to communicate directly makes the coordinator unnecessary and limits the total load on any one thread.

How it works: Under MPLM, a model and its iterations can write commands that start threads, send results to specific threads, wait for replies, or stop threads. The authors built programs that (i) used these commands to solve puzzles and (ii) produced text traces as they considered possible solutions. They trained Qwen3-0.6B-Base on those traces. The puzzles included examples of 3-SAT (deciding whether a boolean formula can evaluate to true) and Sudoku (filling a square grid with numbers, from 1 to the number of cells in a row, column, and box, so no number repeats in any row, column, or box). The description below applies to solving Sudoku. Solving 3-SAT involved a different process.

• A parent thread tracked the parts of a puzzle that had been solved.
• For each cell in a Sudoku grid, its thread determined the correct number by elimination (since a Sudoku grid comes pre-populated with some numbers and a cell can’t repeat a number already in its row, column, or box). While a thread was undecided, it waited to receive numbers from the threads that managed cells in its row, column, and box. As their numbers arrived, it ruled out numbers until one was left. Once it had settled on a number, it sent the number to the parent and related threads and stopped.
• When every thread had stopped, the parent thread reported the solution.

Results: MPLM solved the puzzles faster and in fewer tokens per thread than two earlier approaches: a single thread and an agentic harness that runs parallel threads but routes their results through a coordinator.

• Solving Sudoku grids from 4×4 to 25×25, MPLM was significantly faster on average. For instance, given 9x9 grids, MPLM solved 100 percent in roughly 15 seconds, whereas the parallel method solved 93 percent in roughly 60 seconds. Furthermore, MPLM’s tokens per thread grew more slowly as the grids scaled up. MPLM solved 72 percent of the 25×25 puzzles, where the other two methods reached limits of context or compute imposed by the authors before they learned to solve the problems.
• Solving 3-SAT problems involving formulas that included 8 to 20 variables, MPLM’s accuracy was about even with the parallel method, roughly 92 versus 91 percent. It was marginally faster on average than the parallel method, but it processed some examples much faster — sometimes 2.5x faster — because once the model found a solution in one thread, it could stop the others. The single-thread method did not run beyond 12 variables, having filled the model’s context window.

Yes, but: MPLM's efficiency depends on knowing in advance which threads must communicate with which. It shines where that communication pattern is fixed and easy to work out, as it is in Sudoku and 3-SAT. For open-ended problems, the authors note that finding the pattern may take careful prompting or extra training. Moreover, the Sudoku examples were limited to puzzles that could be solved by elimination (known as naked singles), so the model never had to guess and check or reason more extensively across threads.

Why it matters: Models that perform tasks in a single thread can fill their context windows before they reach a solution. MPLM enables models to distribute the work among many threads that, collectively, can get the job done.

We’re thinking: The authors also prompted two larger models, Qwen3-30B-A3B and Qwen3.6-35B-A3B, to use their method to tackle problems on the LongBench-v2 long-context reasoning benchmark. Both models showed improved accuracy and faster responses (roughly a 2x reduction in average latency) in the MPLM harness. This suggests this method is generalizable to reasoning challenges beyond the relatively simple Sudoku and 3-SAT examples, although its effects do seem to be stronger on smaller models.