activity
20242026
collaborators

7 papers

cs.LG2026

Subliminal Learning as Trait-Direction Drift: A Mechanism and Targeted Control under SFT Distillation

Zhixuan Liu, Zhichen Dong, Yuyu Fan +2

Beyond intended capabilities, model distillation can transfer hidden traits from a teacher. A teacher biased by a system prompt can generate semantically clean training data, such…

cs.LG2026

Scaling Model-Generated Distillation Data Can Make Latent Teacher Traits More Recoverable

Zhichen Dong, Zhixuan Liu, Yuyu Fan +3

Scaling model-generated data is usually viewed as improving distillation: more examples should increase coverage, reduce noise, and produce stronger students. We show a second effe…

cs.CL2026

Decoupled Contrastive Decoding via Expert-Aligned Drafting

Zhixuan Liu, Zhichen Dong, Yuanfu Wang +1

Contrastive Decoding (CD) improves generation quality, but its amateur-model pass makes decoding expensive. Accelerating CD with speculative decoding raises a proposal-alignment qu…

cs.LG2026

Native Reasoning Models: Training Language Models to Reason on Unverifiable Data

Yuanfu Wang, Zhixuan Liu, Xiangtian Li +2

The prevailing paradigm for training large reasoning models--combining Supervised Fine-Tuning (SFT) with Reinforcement Learning with Verifiable Rewards (RLVR)--is fundamentally con…

cs.CL2025

Emergent Response Planning in LLMs

Zhichen Dong, Zhanhui Zhou, Zhixuan Liu +2

In this work, we argue that large language models (LLMs), though trained to predict only the next token, exhibit emergent planning behaviors: $\textbf{their hidden representations…

cs.CL2024

Inference-Time Language Model Alignment via Integrated Value Guidance

Zhixuan Liu, Zhanhui Zhou, Yuanfu Wang +2

Large language models are typically fine-tuned to align with human preferences, but tuning large models is computationally intensive and complex. In this work, we introduce $\texti…