Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
Final Checkpoints Are Not Enough: Analyzing Latent Reasoning Faithfulness Along Training Trajectories
Hengyu Jin, Shu Yang, Di Wang
Latent reasoning methods perform multi-step inference entirely in the model's continuous hidden states, promising more compact and efficient reasoning. However, these opaque hidden…
cs.LG2026
Global Evolutionary Steering: Refining Activation Steering Control via Cross-Layer Consistency
Xinyan Jiang, Wenjing Yu, Di Wang +1
Activation engineering enables precise control over Large Language Models (LLMs) without the computational cost of fine-tuning. However, existing methods deriving vectors from stat…