Transformer训练轨迹决定电路可移除性
Training Trajectories Determine Circuit Removability in Annealable Soft-Prior Transformers
这篇论文揭示了Transformer中小型离散检索任务中电路可移除性取决于训练轨迹而非最终架构，对模型优化有重要启示。
研究人员测试了可退火软先验Transformer模型，发现其在联想回忆任务中，未强制模型在先验激活时表现良好(0.772±0.020)，但在零门设置下性能崩溃(0.095±0.009)。平滑渐变至零训练保留了高零门准确率(0.734±0.028)，而强制零训练、硬切换和事后延续未能恢复相同效果。在马尔可夫归纳任务上也观察到类似模式。
Training Trajectories Determine Circuit Removability in Annealable Soft-Prior Transformers
Soft positional priors can help small Transformers learn retrieval circuits, but it is unclear whether the resulting circuits remain functional once the prior is removed. We test this with an annealable soft-prior Transformer whose attention biases can be learned, faded, or zeroed during training and evaluation. On associative recall, unforced models perform well with the prior active ($0.772 \pm 0.020$) but collapse at zero gate ($0.095 \pm 0.009$). Smooth fade-to-zero training preserves high zero-gate accuracy ($0.734 \pm 0.028$), whereas forced-zero training, hard switching, and post hoc continuation fail to recover the same effect. The pattern also appears on Markov induction. Linear regression ICL provides a boundary case because zero-gate training can learn that task directly. Mechanistic traces show that circuit consolidation occurs after the gate reaches zero, even though the responsible heads vary across seeds. These results suggest that circuit removability in small discrete retrieval tasks depends on the training trajectory, not just the final architecture.
