arXiv:2607. 20457v1 Announce Type: cross Abstract: Inference with large language models (LLMs) on long sequences is computationally expensive due to the quadratic complexity of self-attention.
arXiv:2607. 21752v1 Announce Type: new Abstract: Data-adaptive sparse attention masks substantially outperform fixed patterns (e.
arXiv:2606. 13392v1 Announce Type: new Abstract: Ultra-long-context capability is becoming indispensable for frontier LLMs: agentic workflows, repository-scale code reasoning, and persistent memory all require the model to jointly attend over hundreds of thousands to millions of tokens, yet the quadratic cost of softmax attention makes this untenable at deployment scale.
arXiv:2607. 09052v1 Announce Type: new Abstract: Block sparse attention is a hardware friendly way to alleviate the key-value (KV) cache read bottleneck in large language models (LLMs).
The paper investigates whether recent attention‑mechanism improvements—specifically gated attention, Kimi K3, Kimi Delta Attention, and Attention Residuals—effectively eliminate the attention‑sink problem when scaling language models to a one‑million‑token context window. Using a new diagnostic suite called SinkProbe, the authors evaluate sink mass, massive activation, position‑resolved recall, and the recency gap across four small models that vary only in token mixing and depth. Their findings show that the training objective, rather than the architecture, drives the emergence of attention sinks; gating did not replicate its previously reported benefits at the larger scale, and sink mass, activations, and positional bias behaved independently.
RBS-Attention introduces a training‑free, radius‑bounded sparse prefill strategy for long‑context large language models, addressing the mean dilution problem where a block centroid can miss highly relevant tokens. The method employs two complementary selection branches: a centroid base branch that captures average relevance and a rescue branch that uses the maximum key‑block radius to flag under‑estimated blocks. Experiments on Qwen3 models demonstrate significant speedups—over 20× in standalone prefill‑attention and nearly 6× in end‑to‑first‑token time—while maintaining competitive accuracy compared to dense attention.
CompKV introduces a compensation‑aware sparse attention framework for long‑context LLM inference. It partitions tokens into blocks and optimizes token selection to minimize the error introduced by block‑level mean compensation, using compact block‑level statistics. Experiments on RULER and LongBench‑Pro show CompKV outperforms other sparse baselines and achieves up to a 6.85× speedup over full attention.
HISA: Efficient Hierarchical Indexing for Fine-Grained Sparse Attention proposes a two-stage hierarchical indexer that replaces the flat token scan used in token-level sparse attention mechanisms like DeepSeek Sparse Attention. The method first performs block-level coarse filtering to discard irrelevant regions, then applies the original token-level indexer only within the retained candidate blocks, preserving the same top-sparse pattern for downstream attention. Benchmarks show HISA achieves significant speedups at 64K context and matches the quality of DeepSeek-V3.2 and GLM-5 without additional training.
arXiv:2509. 10406v4 Announce Type: replace Abstract: Pretraining transformers on long sequences (entire code repositories, collections of related documents) is bottlenecked by quadratic attention costs.
Small language models are usually built like large ones and then squeezed onto a CPU afterwards. We did the opposite: we fixed the target first, one user, one token at a time, 4-bit weights, ordinary CPU, and chose the architecture to suit it.
arXiv:2608. 20210v1 Announce Type: cross Abstract: Small language models are usually built like large ones and then squeezed onto a CPU afterwards.
arXiv:2601. 03043v4 Announce Type: replace-cross Abstract: Large language models (LLMs) demonstrate strong capabilities across a wide range of complex tasks and are increasingly deployed at scale, placing significant demands on inference efficiency.