Sparse coverage.

Statistically Grounded Sparse-Feature Interventions for Activation-Space Control in Large Language Models

arXiv:2607.19364v1 Announce Type: new Abstract: Activation steering offers a lightweight alternative to fine-tuning for behavioral control of large language models, but SAE-based s…

LISA: Linear-Indexed Sparse Attention for Efficient Long-Context Reasoning

arXiv:2607.19358v1 Announce Type: new Abstract: Recent advances in long chain-of-thought reasoning models such as DeepSeek-R1 have led to increasingly longer inference context leng…

Iran oil tankers turned back by US blockade, Hormuz traffic sparse

LONDON — Six tankers loaded with Iranian oil have been forced back to Iran by the U.S. blockade in recent days, ship-tracking data shows, underscor...…

Accelerating GPU Inference of Large Language Models with Moderately Unstructured Sparse Weight Matrices

With the growing deployment of large language models (LLMs), LLM inference cost has become a key challenge. Pruning techniques that introduce sparsity into weight matrices can acce…

AlphaZero in Sparsely Rewarded Games: Limits and Auxiliary Supervision

AlphaZero has demonstrated that a neural-guided Monte Carlo Tree Search can achieve superhuman performance, but strong play does not necessarily imply perfect play. We study this g…

A look at the Seckinger school cluster in Georgia, including the US' "first AI-themed educational institution", as parents say AI integration is often sparse (New York Times)

New York Times : A look at the Seckinger school cluster in Georgia, including the US' “first AI-themed educational institution”, as parents say AI integration is often sparse — It …

Learning to Skip Blocks: Self-Discovered Ultrametric Routing for Hardware-Accelerated Sparse Attention

Opus 4.8 Killer: NexusCortex Isn't an LLM – It's a Sparse AI Cortex Built in Go

Experimental sparse cognitive architecture written in Go. SDR attention, ternary compute, memory systems, sleep consolidation, 137 tests. - Branches · office233/Nexuscortex…

MiniMax teases upcoming M3 model with new sparse attention mechanism and 15.6X long-context response speed boost

RAG - Sparse Embedding

Sparse means thinly spread, scattered, or not dense. In sparse embeddings, chunks are converted into...…

Sparse Autoencoders Reveal Cortical Brain-LLM Semantic Mapping

A preprint submitted to arXiv (arXiv:2605.23035) by Dongxin Guo and colleagues presents a mechanistic interpretability approach connecting large language model representations to h…

DemoEvolve: Overcoming Sparse Feedback in Agentic Harness Evolution with Demonstrations

Agent harness evolution improves frozen language-model agents by modifying the executable structures around them. We study this paradigm as a form of sample-efficient fast adaptati…

Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps

Sparse Autoencoders Map Brain-LLM Alignment onto Cortical Semantic Topography

Intermediate layers of large language models (LLMs) best predict human brain responses to language, one of the most robust findings in computational neurolinguistics, yet why remai…

SimInsert: Seamless Video Object Insertion via Regional Sparse Attention Fusion

Video object insertion requires ensuring spatio-temporal coherence and interactive realism, extending far beyond simple content placement. However, current approaches are often hin…

Sparse Compositional Flow Matching by geometric assembly from motion primitives

Embodied trajectories, such as the executable motion sequences of robotic manipulators, underwater vehicles, and mobile robots, are a fundamental output of embodied AI. Modern gene…

SSV: Sparse Speculative Verification for Efficient LLM Inference

Speculative decoding and dynamic sparse attention are two complementary approaches for accelerating long-context LLM inference: the former amortizes target-model execution across m…

Cohere releases Command A+, a sparse MoE open model built for agentic tasks, with 218B total and 25B active parameters, its first under the Apache 2.0 license (Carl Franzen/VentureBeat)

Carl Franzen / VentureBeat : Cohere releases Command A+, a sparse MoE open model built for agentic tasks, with 218B total and 25B active parameters, its first under the Apache 2.0 …

From Sparsity to Simplicity: Enabling Simpler Sequential Replacements via Sparse Attention Distillation

Self-attention serves as the core foundation of large-scale transformer pretraining, but its quadratic token interaction cost makes inference expensive. Replacing attention with si…

FLUIDSPLAT: Reconstructing Physical Fields from Sparse Sensors via Gaussian Primitives

Reconstructing continuous flow fields from sparse surface-mounted sensors is central to aerodynamic design, flow control, and digital-twin instrumentation. Existing neural methods …

Sparse Federated Representation Learning for precision oncology clinical workflows during mission-critical recovery windows

My exploration of this topic began during a late-night research session in early 2024, where I was studying the intersection of federated learning and oncology. I had just read a p…

Surface-Form Neural Sparse Retrieval: Robust Fuzzy Matching for Industrial Music Search

Music search at the scale of Amazon Music presents a unique challenge: queries frequently deviate from indexed metadata due to misspellings, transpositions, and phonetic variations…