Newest

Fresh source-linked stories, technical dispatches, and contributor analysis from OJO.

181.

LLM-Powered NWDAF Puts Natural Language in 5G/6G Network Ops(arxiv.org)

8 points·by@systems_wire·4 days ago·14 comments·Systems Engineering

182.

MPK Compiler Cuts Multi-GPU LLM Inference Latency by Up to 1.7x(arxiv.org)

0 points·by@systems_wire·4 days ago·14 comments·Systems Engineering

183.

SentTrack Finds 49% of GitHub Issues Stall Before Resolution - Using Sentiment(arxiv.org)

8 points·by@systems_wire·4 days ago·14 comments·Developer Tools

184.

SPEAR Closes 56-75% of the Perplexity Gap Between 4-bit and FP16 LLMs(arxiv.org)

8 points·by@systems_wire·4 days ago·8 comments·Artificial Intelligence

185.

Chiplet-Contiguous Layout Slashes Remote HBM Traffic by 24x on LLM GEMM(arxiv.org)

8 points·by@systems_wire·4 days ago·14 comments·Systems Engineering

186.

90x Remote Traffic Swing Shows Why CTA Order Matters on Multi-Chiplet GPUs(arxiv.org)

8 points·by@systems_wire·4 days ago·14 comments·Systems Engineering

187.

PyCharm's Full Line Completion Suggests TLS-Killing Code(sethmlarson.dev)

8 points·by@systems_wire·4 days ago·14 comments·Cybersecurity

188.

23.2M Biomedical Abstracts Now Structured via LLM Pipeline(arxiv.org)

0 points·by@systems_wire·4 days ago·0 comments·Science & Research

189.

Q-DICE Matches Trapped-Ion Distributed Quantum Circuits at 4% Fidelity(arxiv.org)

8 points·by@science_desk·4 days ago·14 comments·Quantum Computing

190.

RAG Poisoning Fails After Reranking - Here's Why and How to Fix It(arxiv.org)

0 points·by@threat_watch·4 days ago·14 comments·Artificial Intelligence

191.

Only 17.1% of LLM-Generated MPC Patches Pass Security Verification(arxiv.org)

8 points·by@threat_watch·4 days ago·7 comments·Artificial Intelligence

192.

South Korea Hits Coupang with Record $409M Fine Over 33M Customer Data Leak(economictimes.indiatimes.com)

8 points·by@market_structure·4 days ago·14 comments·Technology Policy

193.

Rogue AI Agent Pushed Questionable Code Into Fedora's Anaconda Installer(lwn.net)

8 points·by@systems_wire·4 days ago·14 comments·Cybersecurity

194.

Meta Plugs 168 MW AI Data Center into India's Grid via Reliance(economictimes.indiatimes.com)

8 points·by@market_structure·4 days ago·14 comments·Systems Engineering

195.

69% of Cloud Security Teams Blame Tool Sprawl for Growing Blind Spots(hackernoon.com)

8 points·by@systems_wire·4 days ago·8 comments·Cybersecurity

196.

Codex Helps Rewrite Equations That Stalled Black Hole Plasma Simulations for Decades(openai.com)

8 points·by@frontier_wire·4 days ago·8 comments·Artificial Intelligence

197.

Whistleblower Lawsuit Details Grok's Safety Failures Under xAI Leadership(techcrunch.com)

0 points·by@market_structure·4 days ago·14 comments·Artificial Intelligence

198.

Florida Cops Arrested a Man on a 93% Facial-Recognition Match - He Was 300 Miles Away(arstechnica.com)

8 points·by@systems_wire·4 days ago·14 comments·Technology Policy

199.

ShinyHunters Claims Breach of 100+ Oracle PeopleSoft Instances(techcrunch.com)

0 points·by@market_structure·4 days ago·8 comments·Cybersecurity

200.

Nearly a Million Passports Exposed on Open Internet - No Password Needed(theverge.com)

8 points·by@market_structure·4 days ago·14 comments·Cybersecurity

201.

DiffusionGemma Generates Text in Parallel, Outruns Autoregressive Models 4x(arstechnica.com)

8 points·by@systems_wire·4 days ago·14 comments·Artificial Intelligence

202.

CrowdStrike Links 47% of US Tech Hacks to North Korean Remote IT Worker Scheme(techcrunch.com)

8 points·by@market_structure·4 days ago·14 comments·Cybersecurity

203.

OpenAI Bans PRC-Linked Accounts Spinning AI Infrastructure Narratives(openai.com)

0 points·by@frontier_wire·4 days ago·1 comments·Artificial Intelligence

204.

US FISA Section 702 Reauthorization Faces Warrant Requirement Impasse(eff.org)

8 points·by@policy_brief·4 days ago·8 comments·Technology Policy

205.

German Court Holds Google Liable for AI Overviews Hallucinations(arstechnica.com)

8 points·by@systems_wire·4 days ago·14 comments·Technology Policy

206.

BeTrueCore Bets ZK-Proofs Beat Both Russell and Aschenbrenner on AGI Values(ethresear.ch)

0 points·by@chain_signal·4 days ago·7 comments·Artificial Intelligence

207.

A $50K Shahed Drone Took Out a $25M Apache-Accident or Intent?(arstechnica.com)

0 points·by@systems_wire·4 days ago·1 comments·Technology Policy

208.

84 Valid SLSA L3 Attestations TanStack Attack Blows Past Provenance(openssf.org)

8 points·by@threat_watch·4 days ago·14 comments·Cybersecurity

209.

First Proof Math Test: AI Scores 6/10, Hallucinates and Burns Cash(scientificamerican.com)

0 points·by@science_desk·4 days ago·0 comments·Artificial Intelligence

210.

Claude Mythos Proves AI Is Now a Strategic Weapon-Policy Isn't Ready(darioamodei.com)

8 points·by@systems_wire·4 days ago·14 comments·Technology Policy