Quantization-aware training (QAT) aims to recover reasoning capabilities in low-bit large language models (LLMs). Existing recipes typically supervise the quantized student with the original full-precision model on fixed response sequences. However, the resulting low-bit quantized models still struggle on long-horizon...
Bo-Yuan Zheng, Jing Xiong, Chen-Wei Wu et al.· 0 citations
Diffusion language models apply bidirectional attention over the whole context at every denoising step, so a longer context makes each step substantially more expensive. Sparsifying only at inference leaves the pretrained weights unchanged, so the model is never trained under the sparse pattern it runs with. We propose...
Qi Han, Jing Xiong, Zi-Yuan Zhang et al.· 0 citations
On-policy Verbal Distillation is introduced, a framework that uses verbal scores from black-box teachers to rank student-generated sub-trajectories, retaining high-scoring ones and replacing low-scoring ones with teacher-generated continuations and suggests that retaining student-generated prefixes helps preserve explo...
Jing Xiong, Hui Shen, Shansan Gong et al.· arXiv.org· 8 citations
SPS is a training-free latent steering framework that constructs a state-conditioned Direction Bank containing multiple progress-guided steering vectors for different prefix-state regions and applies it at high-uncertainty transitions to guide the next reasoning step toward meaningful progress.
Hengyuan Zhang, Chenming Shang, Zun-Hai Su et al.· 0 citations
PerSyn (Personalized data Synthesis), a novel synthesis strategy that operates under a new ``Route then Generate''paradigm to create data tailored to each student model, enabling it to learn more effectively, is proposed.
Hengyuan Zhang, Shi-Ping Yang, Xiao Liang et al.· Annual Meeting of the Associ...· 21 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.