FedReason：用 Claude Opus 5.5 让 LLM 自动选择联邦学习聚合规则
FedReason: Large Language Models as Weight-Space Meta-Aggregators for Heterogeneous Federated Learning
给联邦学习开发者看的新玩法：Claude Opus 5.5 每 10 轮看一次客户端权重，自己写聚合代码，168 次实验显示比 FedAvg 高 1.26 个点，一次才花 0.79 美元。
FedReason 提出让 LLM 充当联邦学习服务端的元聚合器：每 10 轮调用一次，在沙盒中分析客户端权重，可在 500 张图像的代理集上评分候选方案，并生成一段 Python 聚合程序。在 CIFAR-10 和 CIFAR-100 上用 ResNet-18 共做了 168 次实验，对比 14 个基线，标签偏斜场景下平均比 FedAvg 高 1.26 个百分点。10 个客户端中 3 个作恶时，它从第一轮起就排除所有符号翻转客户端，且不训练根模型时只比 FLTrust 低约 1 个点。60 轮实验的 API 费用平均为 0.79 美元，但去掉代理反馈后准确率会从 73.8% 跌到 38.8%。
FedReason: Large Language Models as Weight-Space Meta-Aggregators for Heterogeneous Federated Learning
Federated learning servers combine client models with a fixed aggregation rule, and each rule in common use is designed for one kind of heterogeneity. We study whether a large language model (LLM) that inspects the client models can choose the rule itself, without any training on the server. In FedReason, the LLM is called every ten rounds. It analyses the weights by running code in a restricted server-side sandbox, sees only the printed results, may score a few candidates on a 500-image proxy set, and submits a short Python aggregation program. The server applies the program until the next call and falls back to FedAvg whenever its output is invalid. Full weight tensors never leave the server. We evaluated FedReason with Claude Opus 5.5 on CIFAR-10 and CIFAR-100 with ResNet-18 under Dirichlet label skew, width-heterogeneous sub-models and Byzantine clients, in 168 runs against 14 baselines. Across 12 paired label-skew runs it beat FedAvg by 1.26 percentage points on average (10 wins), and a control that spends the same per-round proxy budget on choosing among standard rules by 1.08 points. It was the strongest server-side rule on CIFAR-100 but did not beat SCAFFOLD, which changes client training. With three of ten clients malicious, it excluded all sign-flipping clients from the first round and stayed about one point below FLTrust without training a root model. Without proxy feedback, accuracy fell from 73.8% to 38.8% in one seed. A 60-round run cost 0.79 USD in API fees on average.