ConceptTTT / Test Time Training
Test-Time Training
Test-Time Training (TTT) is a machine learning paradigm that permits lightweight parameter updates or adaptations during the model inference phase, dissolving the conventional boundary between training and inference. This allows a model to adaptively adjust based on the current input at test time, making it effective at addressing distribution shift problems.
Timeline (last 90 days)
Oct 12
在搜索与训练耦合的循环中，仅保存最新权重无法完整恢复实验状态，因为模型检查点与搜索档案共同描述同一个演化中的实验，恢复时必须成对还原。
Unverified50%
Oct 12
在 TTT 与搜索耦合的场景中，PUCT 搜索不是一次性完成的，而会在每个训练 step 后继续扩展，因此搜索树本身（哪些节点被访问、访问次数、价值估计等）构成必须被持久化的状态。
Unverified50%
Jul 17
测试时训练（TTT）在推理阶段利用当前输入的自监督信号对部分参数进行短暂局部梯度更新再执行预测
Unverified50%