Attention-state memory externalizes long prefixes into a lightweight lookup table of precomputed attention states, yielding higher accuracy than standard in-context learning at fixed memory budgets and lower latency than full attention.
Flashattention-3: Fast and accurate attention with asynchrony and low-precision.Advances in Neural Information Processing Systems, 37:68658–68685
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 1
method 1
other 1
citation-polarity summary
representative citing papers
OSCAR achieves near-BF16 accuracy for 2-bit KV cache quantization by using offline spectral covariance-aware rotations aligned with attention, plus a custom deployable INT2 kernel compatible with paged serving.
SOMA estimates a local response manifold from early turns and adapts a small surrogate model via divergence-maximizing prompts and localized LoRA fine-tuning for efficient multi-turn serving.
Kling-Omni is a unified multimodal generative system that produces cinematic videos from diverse inputs by integrating generation, editing, and intelligent reasoning in a single end-to-end model.
A 6B single-stream diffusion transformer trained with heavily curated data reaches top open-source image-generation quality in 314K H800 GPU hours, releasing Turbo and Edit variants.
CompactAttention accelerates chunked-prefill attention via Block-Union KV Selection, delivering up to 2.72x speedup at 128K context on LLaMA-3.1-8B while matching dense accuracy on RULER.
OmniRefine introduces alignment-aware chunk refinement via similarity and dynamic programming followed by modality-cooperative token compression, achieving near-baseline accuracy at 44% token retention on WorldSense.
citing papers explorer
-
Context Memorization for Efficient Long Context Generation
Attention-state memory externalizes long prefixes into a lightweight lookup table of precomputed attention states, yielding higher accuracy than standard in-context learning at fixed memory budgets and lower latency than full attention.
-
OSCAR: Offline Spectral Covariance-Aware Rotation for 2-bit KV Cache Quantization
OSCAR achieves near-BF16 accuracy for 2-bit KV cache quantization by using offline spectral covariance-aware rotations aligned with attention, plus a custom deployable INT2 kernel compatible with paged serving.
-
SOMA: Efficient Multi-turn LLM Serving via Small Language Model
SOMA estimates a local response manifold from early turns and adapts a small surrogate model via divergence-maximizing prompts and localized LoRA fine-tuning for efficient multi-turn serving.
-
Kling-Omni Technical Report
Kling-Omni is a unified multimodal generative system that produces cinematic videos from diverse inputs by integrating generation, editing, and intelligent reasoning in a single end-to-end model.
-
Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer
A 6B single-stream diffusion transformer trained with heavily curated data reaches top open-source image-generation quality in 314K H800 GPU hours, releasing Turbo and Edit variants.
-
CompactAttention: Accelerating Chunked Prefill with Block-Union KV Selection
CompactAttention accelerates chunked-prefill attention via Block-Union KV Selection, delivering up to 2.72x speedup at 128K context on LLaMA-3.1-8B while matching dense accuracy on RULER.
-
OmniRefine: Alignment-Aware Cooperative Compression for Efficient Omnimodal Large Language Models
OmniRefine introduces alignment-aware chunk refinement via similarity and dynamic programming followed by modality-cooperative token compression, achieving near-baseline accuracy at 44% token retention on WorldSense.