Evaluating the accuracy of KV cache reuse techniques
Study reveals KV cache reuse evaluation metrics artificially inflate effectiveness; proposes faithful accuracy measurement methodology for retrieval-augmented generation.
Search the full wire by company, model, lab, or keyword. Every story we have ever aggregated.
Study reveals KV cache reuse evaluation metrics artificially inflate effectiveness; proposes faithful accuracy measurement methodology for retrieval-augmented generation.
VLMs fail on multi-layer text images; DecoyBench dataset exposes typographic attack vulnerability across six model families.
AFA-Net applies differential attention to EEG-based auditory speaker detection, achieving 96.8% accuracy on noisy signals.
Probe-guided steering effectiveness increases across Pythia pretraining/post-training checkpoints; in-context state decodability rises during training.
EmoSpeechBrain fuses EEG and speech for emotion recognition using differential attention to suppress noise before shared representation learning.
Intent2Tc uses LLMs to translate business-level QoS intents into validated Linux traffic-control policies via closed-loop framework.
ActKV prioritizes KV cache entries by action-generation contribution to reduce agent inference memory overhead and boost throughput.
End-to-end driving models suffer open-loop/closed-loop training mismatch; endpoint-constrained trajectory optimization improves physical coherence.
Highlight-Then-Summarize trains LLMs to compress sparse evidence before reasoning on long contexts; H2S-Dataset covers 11 benchmark families.
ReVerPi case study: context projection with archived observations shows equal success on 12/15 completed pairs vs. full continuation.
At TechCrunch Disrupt 2026, Ricursive Intelligence co-founders Anna Goldie and Azalia Mirhoseini will take the Disrupt Stage to discuss closing the loop between AI and chip development Save up to $200 on your pass before today ends.
Your next opportunity could be one conversation away. Get your Expo+ Pass for just $75. Limited to the first 100 qualifying people.
Last 24 hours to save up to $200 on your TechCrunch Disrupt 2026 pass. Leave the event further in your startup's trajectory than where you started. Don't miss your chance to save and to push the needle.
I used Google’s Nest Doorbell (left), Aqara’s HomeKit-compatible G400 (center), and the Ring Pro 4K (right) to test the new AI features of Gemini for Home, Apple Intelligence for Home, and Amazon’s Ring. | Photo by Jennifer Pattison Tuohy / The Verge A few years back, I was at a beachside Easter egg hunt, watching my kids dash through sand dunes searching for sweet treats. My phone buzzed in my pocket; I ignored it. A moment later, it buzzed again. I pulled it out, glanced down, and saw a "motion detected" notification from my security camera. "Just my dog in the yard," I thought, dismissing ...
After teasing its new Copilot "super app" last month, Microsoft is officially unveiling it today. The redesigned Copilot app bundles three AI capabilities into a single interface of chat, coding, and agents. As part of the launch, Microsoft is also rebranding Scout, the AI personal assistant it unveiled at Build earlier this year, as Autopilot. This new Copilot app will have three tabs for Home, Code, and Autopilot, with Home combining Copilot Chat and Cowork together. It's the default landing experience for Copilot users, and Microsoft is also planning to include a Today feature inside it th...
Vendors rolling out AI have an "incentive to deny as many claims as possible."
The US government wants to spend $30.3 million over the next five years on an improved form of lie detector, according to a Department of Defense budget request. The program, called “Polygraph+” or “Polygraph Next”, will focus on scoring algorithms that use artificial intelligence and machine learning and on a technique called “standoff sensing”, which…
Latent Space newsletter announces operational status with vague reference to behind-the-scenes work; no specific technical content or announcements.
The Silicon Valley firm is aligning its India fundraising cycle with its global funds for the first time, as it shifts to a shorter investment period.
Runway's GWM Worlds 2 uses persistent context and timed actions to generate real-time video and audio via world models.
Cohere launches Compass as cloud SaaS in private beta for enterprise search applications.
Simon Willison argues coding agents increase software engineering complexity without proper discipline and domain expertise.
commit-rewriter 0.2 adds support for non-default Git branches via --branch flag.
Google's new Gemini 3.8 Live update lets users have conversations with the model while watching an animated AI persona respond in real time. The "Live Avatar" will lip-sync and show different facial expressions during conversations, but it's currently only available to Gemini Enterprise customers. As noted by Google, Live Avatar can transition between the 97 languages it supports "without degrading video fidelity or introducing visual drift." One video shared by Google shows its Live Avatar talking in English and Japanese, with its mouth animation lining up with what it's saying in both langu...
Buy one pass to TechCrunch Disrupt 2026 and get 50% off a second of the same ticket type. Register before event starts on October 13 at 8 a.m. PT.
Prism's larger goal is open-weight AI that runs on devices and makes better use of the computing power they already have.
Million-dollar fines won’t end billionaires’ data center pollution, neighbors fear.
The notice would allow Oracle to delay payments should the facility miss its 2028 target to come online.