Tag
This paper introduces Semantic Step Prediction, which applies geometric regularization at reasoning step boundaries rather than random token positions, achieving 168× better multi-step latent forecasting on ProcessBench compared to frozen baselines.
This paper investigates failures of final-token safety probes on jailbreak prompts, finding that harmful content can be distributed across earlier tokens and missed by the final readout. It proposes a PCA-HMM trajectory model as a diagnostic tool that recovers many misses without the false positives of naive token pooling.