2025.10.23 | 线性注意力显存降十倍；动态裁剪PPO稳提分
"长上下文推理" 相关文章
2025.10.23 | 线性注意力显存降十倍；动态裁剪PPO稳提分
Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning
OWL: Overcoming Window Length-Dependence in Speculative Decoding for Long-Context Inputs
8GB显卡的逆袭！SSD换显存，3060 Ti硬跑100k长上下文
NVIDIA Rubin CPX Accelerates Inference Performance and Efficiency for 1M+ Token Context Workloads
推理算力“正在耗尽”？百万Token时代开启新一轮竞逐
英伟达下一代GPU登场，Rubin CPX一次推理数百万Token，网友：这是头野兽
SparK: Query-Aware Unstructured Sparsity with Recoverable KV Cache Channel Pruning
Michelangelo: An Artificial Intelligence Framework for Evaluating Long-Context Reasoning in Large Language Models Beyond Simple Retrieval Tasks
Fact or Fiction? NOCHA: A New Benchmark for Evaluating Long-Context Reasoning in LLMs
