N, Long:
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
Restartable Sequences
May 31st, 2026 @ justine's web page The best kept secret at the frontier of system programming right now is the Linux 4.18+ (c. 2018) concept of restartable sequences or rseq for short. They allow you to create thread-safe data structures without locks or atomics which scale to microprocessors with many cores. It's currently only possible to use rseq on Linux using handwritten assembly code.
Towards Long-Horizon Interpretability: Efficient and Faithful Multi-Token Attribution for Reasoning LLMs
arXiv:2602.01914v4 Announce Type: replace Abstract: Token attribution methods provide intuitive explanations for language model outputs by identifying causally important input tokens. However, as modern LLMs increasingly rely on extended reasoning chains, existing schemes face two critical challenges: (1) efficiency bottleneck, where attributing a target span of M tokens within a context of length N requires O(M*N) operations, making long-context attribution prohibitively slow; and (2)...
Repurposing drugs to treat Amoebic Gill Disease in Atlantic Salmon
Neoparameoba perurans causes Amoebic Gill Disease (AGD), a major parasitic disease of marine-phase Atlantic salmon and rainbow trout worldwide. Treatment options are limited to freshwater baths, which are costly at scale and exhibit only limited long-term efficacy. N. perurans contains an obligate eukaryotic symbiont, Perkinsela-like organism (PLO).
Stochastic Sparse Attention for Memory-Bound Inference
arXiv:2605.01910v2 Announce Type: replace Abstract: Autoregressive decoding becomes bandwidth-limited at long contexts, as generating each token requires reading all $n_k$ key and value vectors from KV cache. We present Stochastic Additive No-mulT Attention (SANTA), a method that sparsifies value-cache access by sampling $S \ll n_k$ indices from the post-softmax distribution and aggregates only those value rows. This yields an unbiased estimator of the post-softmax value aggregation while...
Mosquitoes learn to link the smell of DEET with a blood meal, new study finds
Mosquitoes learn to link the smell of DEET with a blood meal, new study finds Sadie Harley Scientific Editor Andrew Zinin Lead Editor Mosquito repellents are key to protecting ourselves from mosquito bites and the pathogens they might carry. The most widely used active ingredient in insect repellents is N,N-diethyl-meta-toluamide, commonly known as DEET. Highly effective, long-lasting (approximately five hours) and cheap to make, DEET is a gold-standard insect repellent.
Bhutan experiences 5.8-magnitude earthquake; tremors felt in parts of Northeast India
A magnitude 5.8 earthquake struck Bhutan late on Sunday night, triggering multiple aftershocks and sending tremors across parts of Northeast India, according to the National Centre for Seismology (NCS). The earthquake occurred at 11.06pm IST at a depth of 26km, with its epicentre located in Bhutan. The epicentre was about 252km from parts of Assam, and tremors were felt in several areas across India's northeastern region following the quake.
Graph Mamba Survival Analysis Based on Topology-Aware ordering
arXiv:2606.02602v1 Announce Type: new Abstract: In computational pathology, Whole Slide Images (WSIs) survival analysis is crucial for patient prognosis assessment, but it faces multiple technical challenges. Although the Transformer captures long-range dependencies through its self-attention mechanism, its $O(N^2)$ time complexity causes a severe computational bottleneck in large-scale WSIs graph structures. The Mamba model breaks through the Transformer's computational bottleneck with...
Test-case reducers are underappreciated debugging tools
Test-case reducers are less well known than they should be, and those who are aware of them don’t always realise the variety of ways we can use – perhaps even abuse! – them. In this post, I’m going to explore some of the things I’ve learnt while using these wonderful tools. I’ll start at the basics, because the idea is so simple that it can be hard to believe it works.
AURA: Action-Gated Memory for Robot Policies at Constant VRAM
Announce Type: new Abstract: The KV-cache is the right memory for datacenters but the wrong memory for robots. Datacenter inference batches many short requests and resets them, amortizing an attention cache across a crowd. Embodied agents instead run one long, non-resetting episode on bandwidth-limited edge hardware, where high-bandwidth memory and flash are scarce, flash has finite write endurance, and memory writes rather than compute can become the binding constraint.
Transcriptionally Active HIV Reservoirs Enriched for Interferon-Inducible APOBEC3-Related Mutational Signatures Associate with Reduced Neuroaxonal Integrity
Despite sustained viral suppression with effective antiretroviral therapy (ART), neuroaxonal integrity persist in a subset of people with HIV-1 (PWH). We investigated whether interferon (IFN)-inducible APOBEC3-associated mutational signatures, reflected by premature stop codons (PSCs), and reservoir-associated drug resistance mutations (RS-DRMs) within transcriptionally active HIV-1 reservoirs influence neuroaxonal integrity in fully virally suppressed individuals. Peripheral CD4 T cells...