This work introduces DecisionPivot, a diagnostic benchmark comprising 48 tasks and 40,033 items across seven decision roles that provides fine-grained diagnostics of decision failures that aggregate evaluation alone cannot reveal, and evaluates 40 decision models and 14 general models using documented, model-specific p...
Xia-Lie Zhuang, Ming-Cong Lei, Zun-Hai Su et al.· 1 citation
This work introduces ReaDy, the first speculative decoding method, to leverage HLA recurrent states for position-specific context retrieval during drafting, and achieves an average $3.16\times decoding speedup over autoregressive decoding with Qwen3.
Bo Sun, Zun-Hai Su, Xia-Lie Zhuang et al.· 0 citations
Multi-Head Latent Attention (MLA) enables expressive multi-head attention with compact caches for its content and decoupled RoPE paths, yet cache memory still scales linearly with context length and batch size. In this work, we establish a systematic model of MLA's dual-path quantization errors, characterizing their di...
Zun-Hai Su, Yuxuan Sun, Jian-Chao Tan et al.· 0 citations
UniMoMo, a post-training compression framework formulated as a constrained graph coarsening problem, is introduced, and a layer-adaptive protection mechanism that restricts the merging of high-traffic experts based on their routing exposure is introduced.
DAMP uses both quantization-error energy and decay-based persistence to identify high-risk channels during offline calibration and stores these channels at higher precision and the remainder in INT8, the first to study post-training quantization of recurrent states in GDN and KDA based language models.
Tao Zhang, Jian-Chao Tan, Ping-Wei Sun et al.· 6 citations· ⚡3
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.