Boltz-Perturb: The Path Not Taken. Unlocking Generative Diversity in Co-Folding Models via Training-Free Conditioning Perturbation

AlphaFold3-style co-folding models produce low-diversity predictions that frequently miss correct small molecule binding modes. We present Boltz-Perturb, a training-free perturbation framework that injects time-annealed noise into the conditioning signals of Boltz-2's denoising module during inference. True-coordinate injection experiments establish that the model's learned energy landscape already contains correct binding-mode basins, reframing the problem as a sampling deficiency. We introduce Token Bias Perturbation (TBP) and Token Conditioning Perturbation (TCP), targeting attention biases and token embeddings, respectively. Notably, TCP raised the oracle success rate from 17.7\% to 30.6\% using three-fold fewer samples across 62 targets. To our knowledge, this is the first perturbation analysis of an AF3-style co-folding architecture for small-molecule binding mode diversity, demonstrating that conditioning perturbation can unlock latent capacity for drug discovery without retraining.