4 papers
RippleKV: Cross-Layer KV Cache Allocation via Perturbation Propagation
Dongjie Xu, Kai Qian, Julius +6
Long-context LLM inference is bottlenecked by KV cache memory, yet distributing a limited cache budget across layers remains challenging. Existing methods rely on proxies such as l…
Attention Sink in Transformers: A Survey on Utilization, Interpretation, and Mitigation
Zunhai Su, Hengyuan Zhang, Wei Wu +24
As the foundational architecture of modern machine learning, Transformers have driven remarkable progress across diverse AI domains. Despite their transformative impact, a persiste…
Quantization Hurts Reasoning? An Empirical Study on Quantized Reasoning Models
Ruikang Liu, Yuxuan Sun, Manyi Zhang +5
Recent advancements in reasoning language models have demonstrated remarkable performance in complex tasks, but their extended chain-of-thought reasoning process increases inferenc…
FlatQuant: Flatness Matters for LLM Quantization
Yuxuan Sun, Ruikang Liu, Haoli Bai +10
Recently, quantization has been widely used for the compression and acceleration of large language models (LLMs). Due to the outliers in LLMs, it is crucial to flatten weights and…