Dopamine: Brain Modes, Not Brains
This addresses the interpretability challenge in PEFT for researchers, though it's an incremental approach with reduced expressivity limitations.
The paper tackles the problem of interpreting parameter-efficient fine-tuning (PEFT) methods by proposing Dopamine, an activation-space technique that adapts models by selecting and rescaling existing computations rather than modifying weights. As a proof of concept on MNIST variants, it improves rotated accuracy over frozen baselines with only a few hundred trainable parameters per layer, trading some accuracy for better interpretability compared to LoRA.
Parameter-efficient fine-tuning (PEFT) methods such as \lora{} adapt large pretrained models by adding small weight-space updates. While effective, weight deltas are hard to interpret mechanistically, and they do not directly expose \emph{which} internal computations are reused versus bypassed for a new task. We explore an alternative view inspired by neuromodulation: adaptation as a change in \emph{mode} -- selecting and rescaling existing computations -- rather than rewriting the underlying weights. We propose \methodname{}, a simple activation-space PEFT technique that freezes base weights and learns per-neuron \emph{thresholds} and \emph{gains}. During training, a smooth gate decides whether a neuron's activation participates; at inference the gate can be hardened to yield explicit conditional computation and neuron-level attributions. As a proof of concept, we study ``mode specialization'' on MNIST (0$^\circ$) versus rotated MNIST (45$^\circ$). We pretrain a small MLP on a 50/50 mixture (foundation), freeze its weights, and then specialize to the rotated mode using \methodname{}. Across seeds, \methodname{} improves rotated accuracy over the frozen baseline while using only a few hundred trainable parameters per layer, and exhibits partial activation sparsity (a minority of units strongly active). Compared to \lora{}, \methodname{} trades some accuracy for substantially fewer trainable parameters and a more interpretable ``which-neurons-fire'' mechanism. We discuss limitations, including reduced expressivity when the frozen base lacks features needed for the target mode.