Daily archive

Refresh calendar

Pick a day and revisit the stories TechEveryday saved for you.

Artificial Intelligence / Machine Learning · 5

Automation / Agentic Systems · 5

Research Papers · 3

Embedded Systems · 5

Computer Systems · 5

Developer Tools / Open Source · 5

Cloud / Infrastructure · 5

Archived section

Artificial Intelligence / Machine Learning · 5

Artificial Intelligence / Machine Learning
arXiv9/24/2026
Recently

A Living Benchmark for Information Retrieval from Electronic Health Records

Large language model (LLM)-based clinical assistants are increasingly being integrated into electronic health record (EHR) systems, transforming how clinicians retrieve and synthesize information from patient records. Their safety and utility depend on rigorous evaluation, yet existing benchmarks are manually curated, costly to update, and rapidly become obsolete with evolving technological advancements. We present a scalable framework that automatically generates question--answer pairs from longitudinal EHR notes. Authors: Jordan L. Cahoon, Chloe O. Stanwyck, Sulaiman Somani.

Why it matters

Read this for the paper's specific claim in Artificial Intelligence / Machine Learning: Large language model (LLM)-based clinical assistants are increasingly being integrated into electronic health record (EHR) systems, transforming how clinicians retrieve and synthesize information from patient records.

Primary paperarxivcs.AI
Original
Artificial Intelligence / Machine Learning
arXiv9/24/2026
Recently

How Reproducible Are Evaluation Conclusions? A Self-Audit of LLM-Inferred Prompt Structure

Evaluations of LLM systems routinely average over small prompt sets and report models as a ranked table. We ask how much confidence such a table deserves, using LLM-based prompt-structure inference as the case study: eight open model variants across five families and 8B to 675B parameters, caching disabled, 293 raw intermediate representations persisted. The measured phenomenon is unstable to begin with. Authors: Dipankar Sarkar.

Why it matters

Read this for the paper's specific claim in Artificial Intelligence / Machine Learning: Evaluations of LLM systems routinely average over small prompt sets and report models as a ranked table.

Primary paperarxivcs.CL
Original
Artificial Intelligence / Machine Learning
arXiv9/23/2026
Recently

StudentBench: AI and human tutoring yield equivalent GRE learning gains

Artificial intelligence offers an unprecedented opportunity to augment human capabilities, yet progress at the frontier has focused primarily on advancing model capabilities. We introduce StudentBench, a suite of AI teaching evaluations and a public platform that enables large-scale data collection with over 175,000 student-AI messages to study whether large language models (LLMs) produce learning gains equivalent to human tutoring. Using StudentBench, we measured learning gains on Quantitative and Verbal GRE questions across 2,383 human participants receiving AI tutoring, human tutoring, or no tutoring. Authors: Curtis Northcutt, Inaara Hasmani, Kevin Feng.

Why it matters

Read this for the paper's specific claim in Artificial Intelligence / Machine Learning: Artificial intelligence offers an unprecedented opportunity to augment human capabilities, yet progress at the frontier has focused primarily on advancing model capabilities.

Primary paperarxivcs.AI
Original
Artificial Intelligence / Machine Learning
NVIDIA Technical Blog9/24/2026
Recently

Efficient MoE Training for Biological Foundation Models

As language models grow, scaling dense architectures becomes increasingly expensive. In a dense transformer, every token passes through every layer, so adding... As language models grow, scaling dense architectures becomes increasingly expensive.

Why it matters

Read this for the official technical update in Artificial Intelligence / Machine Learning: As language models grow, scaling dense architectures becomes increasingly expensive.

Official source
Original
Artificial Intelligence / Machine Learning
Apple Machine Learning Research9/23/2026
Recently

Compressing Streaming Neural Audio Encoders via Latent-Space Distillation

System-wide Dictation on Apple devices runs entirely on-device, and the speech it transcribes reaches the foundation model through a tokenizer: an encoder that maps short windows of waveform onto the representation the language model reads. Because that model is sparsely activated under Instruction-Following Pruning, only a small subset of its experts occupies DRAM at any time, so the always-on tokenizer competes for the same memory, and its parameter count bears directly on power and latency.

Why it matters

Read this for the official technical update in Artificial Intelligence / Machine Learning: System-wide Dictation on Apple devices runs entirely on-device, and the speech it transcribes reaches the foundation model through a tokenizer: an encoder that maps short windows of waveform onto the representation the language model reads.

Official source
Original

Archived section

Automation / Agentic Systems · 5

Automation / Agentic Systems
arXiv9/24/2026
Recently

GRASP: Generating, Revising, and Assessing for Strategic Planning with Agentic AI

Large Language Models (LLMs) typically exhibit a performance profile where reliability degrades as task complexity increases. We address the challenge of generating high-quality natural language executable plans for complex tasks by introducing GRASP, a strategy-aware, multi-stage planning framework. GRASP decouples the planning pipeline across specialized, context-isolated modules: it pre-compiles global macro-guidelines (GenPlan), explores alternative localized strategies within isolated context windows (RevPlan), and independently evaluates trajectories using a multi-criteria discriminator (VerPlan). Authors: Arunabh Srivastava, Mohammad A., Khojastepour.

Why it matters

Read this for the paper's specific claim in Automation / Agentic Systems: Large Language Models (LLMs) typically exhibit a performance profile where reliability degrades as task complexity increases.

Primary paperarxivcs.AI
Original
Automation / Agentic Systems
arXiv9/24/2026
Recently

Coding Agents for Generalized Task and Motion Planning Problems

Task and motion planning (TAMP) problems remain difficult even with full observability and object-centric states because discrete decisions are tightly coupled to geometric, kinematic, and dynamic constraints. Generalized TAMP addresses this difficulty by exploiting regularities across problem instances to reduce planning effort on new instances. However, existing methods require substantial TAMP-specific engineering. Authors: Matteo Merler, Bowen Li, Josh Roy.

Why it matters

Read this for the paper's specific claim in Automation / Agentic Systems: Task and motion planning (TAMP) problems remain difficult even with full observability and object-centric states because discrete decisions are tightly coupled to geometric, kinematic, and dynamic constraints.

Primary paperarxivcs.RO
Original
Automation / Agentic Systems
arXiv9/24/2026
Recently

Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure

A central concern in AI safety is that agents may treat oversight as an obstacle when it conflicts with completing their goals. We study instrumental evasion, the propensity of LLM agents to circumvent runtime monitoring as a means of completing ordinary tasks. We introduce EvasionBench, a benchmark of 50 diverse task-policy pairs in which completing the task requires an operation prohibited by a runtime monitor. Authors: David Schmotz, Derck Prinzhorn, Luca Beurer-Kellner.

Why it matters

Read this for the paper's specific claim in Automation / Agentic Systems: A central concern in AI safety is that agents may treat oversight as an obstacle when it conflicts with completing their goals.

Primary paperarxivcs.CR
Original
Automation / Agentic Systems
GitHub Blog9/25/2026
Recently

GitHub Copilot app for Beginners: How to build custom workflows with canvases

Describe the interface you need in plain English, then let the agent build a live surface you can both use and update—so you spend less time adapting to tools and more time getting work done. The post GitHub Copilot app for Beginners: How to build custom workflows with canvases appeared first on The GitHub Blog.

Why it matters

Read this for the official technical update in Automation / Agentic Systems: Describe the interface you need in plain English, then let the agent build a live surface you can both use and update—so you spend less time adapting to tools and more time getting work done.

Official sourceAI & MLGitHub Copilot
Original
Automation / Agentic Systems
arXiv9/24/2026
Recently

RAPID: Robot Agentic Programming from Demonstrations

Coding agents have demonstrated enormous success in solving complex programming problems. To leverage their potential for robot systems, this work introduces Robot Agentic Programming from Demonstrations (RAPID), which automatically generates, verifies, and refines robot programs, given a single visual human demonstration. The iterative agentic loop of code refinement requires several key ingredients: (i) a testable task specification, (ii) action primitives for robot execution, and (iii) an interactive environment for program execution and verification. Authors: Yuyao Liu, Jiayuan Mao, David Hsu.

Why it matters

Read this for the paper's specific claim in Automation / Agentic Systems: Coding agents have demonstrated enormous success in solving complex programming problems.

Primary paperarxivcs.RO
Original

Archived section

Research Papers · 3

Research Papers
arXiv9/24/2026
Recently

Beyond Driving: Envisioning Activities in Future Autonomous Vehicles through Experience-Centered Design

Autonomous vehicles (AVs) are poised to fundamentally alter personal transportation, offering occupants the freedom to engage in various non-driving-related activities (NDRAs). However, our current understanding of how people might actually use this time in fully autonomous vehicles (FAVs) is limited. Traditional research methods often struggle to capture the influence of diverse travel circumstances and purposes when exploring future scenarios. Authors: Keqi Chen, Xiao Xue, Xinyi Liu.

Why it matters

Read this for the paper's specific claim in Research Papers: Autonomous vehicles (AVs) are poised to fundamentally alter personal transportation, offering occupants the freedom to engage in various non-driving-related activities (NDRAs).

Primary paperarxivcs.HC
Original
Research Papers
arXiv9/23/2026
Recently

A Gmail-Based Phishing Detection Prototype for Nigerian Fintech Emails Using Sender Checks and BiLSTM Classification

Phishing emails that impersonate Nigerian fintech providers can combine deceptive sender addresses, lookalike links, and locally familiar language. This study presents a Gmail browser extension that integrates sender-domain and URL checks with a bidirectional long short-term memory (BiLSTM) classifier. The extension compares visible sender addresses and links with profiles for eight fintech platforms, obtains a phishing probability from a locally hosted Flask service, and displays a legitimate, warning, or phishing verdict when an email is opened. Authors: Gideon Francis Oghie, Uche Emmanuel Unoke.

Why it matters

Read this for the paper's specific claim in Research Papers: Phishing emails that impersonate Nigerian fintech providers can combine deceptive sender addresses, lookalike links, and locally familiar language.

Primary paperarxivcs.CR
Original
Research Papers
arXiv9/24/2026
Recently

T-Backdoor: Exploiting Temporal Redundancy in Neuromorphic Data for Spike-preserving Backdoor Attacks on SNNs

Backdoor attacks are a serious security threat to deep neural networks (DNNs) and remain largely underexplored for spiking neural networks (SNNs). Existing attacks primarily introduce spatiotemporal triggers that induce deviations in the spike distribution of poisoned samples relative to their clean counterparts. To address this limitation, this work proposes a novel backdoor attack on SNNs, termed T-Backdoor, which operates using purely temporal triggers such as Rate, Latency, and Jitter without introducing any spatial perturbation, making the shift in spike distributions significantly harder to detect. Authors: Abdullah Arafat Miah, Kevin Vu, Yu Bi.

Why it matters

Read this for the paper's specific claim in Research Papers: Backdoor attacks are a serious security threat to deep neural networks (DNNs) and remain largely underexplored for spiking neural networks (SNNs).

Primary paperarxivcs.CR
Original

Archived section

Embedded Systems · 5

Embedded Systems
Hackaday9/26/2026
Recently

You Can Make a Microprocessor That's All Your Own

For a while now we've followed the slow progression of affordable integrated circuit fabrication, and through the likes of Tiny Tapeout we've seen impressive strides made.

Why it matters

Read this for the engineering context in Embedded Systems: For a while now we've followed the slow progression of affordable integrated circuit fabrication, and through the likes of Tiny Tapeout we've seen impressive strides made.

Trusted sourcehardwarechip fabrication
Original
Embedded Systems
arXiv9/24/2026
Recently

Rolling-WAM: World Action Models with Rolling Imagination

World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation. However, completing the joint video-action denoising process at each replanning cycle incurs substantial latency, delaying action updates and limiting closed-loop responsiveness. We present Rolling-WAM, a formulation that distributes joint denoising across successive replanning cycles. Authors: Yinghua Zhou, Junjie Ye, Yiqi Zhao.

Why it matters

Read this for the paper's specific claim in Embedded Systems: World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation.

Primary paperarxivcs.RO
Original
Embedded Systems
arXiv9/24/2026
Recently

ReVAMP: Vector-Accelerated Motion Planning for Kinematically-Constrained Systems via Reparameterization

Robots often must satisfy one or more constraints during motion planning for real-world tasks. When such constraints reduce the valid configuration space to a measure-zero subset, sampling based planning algorithms require modifications to draw feasible samples. For many common end-effector constraints, parameterizations built on inverse kinematics (IK) provide an alternate formulation where the constraints are satisfied by construction, allowing directly sampling the feasible set. Authors: Shrutheesh R. Iyer, Thomas Cohn, Zachary Kingston.

Why it matters

Read this for the paper's specific claim in Embedded Systems: Robots often must satisfy one or more constraints during motion planning for real-world tasks.

Primary paperarxivcs.RO
Original
Embedded Systems
arXiv9/24/2026
Recently

Self-Adaptive VLA for Robust Robot Deployment

While Vision-Language-Action (VLA) models demonstrate impressive capabilities in robotic manipulation, their memoryless nature renders them brittle to test-time environment shifts, particularly hardware shifts caused by wear or imperfect calibration. Enabling these models to self-adapt during deployment without requiring continuous on-site recalibration remains a critical bottleneck for real-world scalability. In this work, we introduce Self-Adaptive VLA, a novel post-training recipe that enables the policy to iteratively adapt to deployment-time hardware shifts leveraging its own rollouts as context. Authors: Hongxin Zhang, Chunru Lin, Tsun-Hsuan Wang.

Why it matters

Read this for the paper's specific claim in Embedded Systems: While Vision-Language-Action (VLA) models demonstrate impressive capabilities in robotic manipulation, their memoryless nature renders them brittle to test-time environment shifts, particularly hardware shifts caused by wear or imperfect calibration.

Primary paperarxivcs.RO
Original
Embedded Systems
Hackster.io9/25/2026
Recently

A Tiny Microcontroller Build Revives Forgotten Computing History

Long before digital computers became the obvious choice for tackling difficult mathematical problems, researchers built dedicated hardware that could outperform them. These specialized machines weren't programmable in the modern sense. Instead, they were designed to solve a narrow class of number theory problems as quickly as possible.

Why it matters

Read this for the engineering context in Embedded Systems: Long before digital computers became the obvious choice for tackling difficult mathematical problems, researchers built dedicated hardware that could outperform them.

Trusted source
Original

Archived section

Computer Systems · 5

Computer Systems
arXiv9/24/2026
Recently

MQSS-Selector: RL-Guided Pass Selection for an MLIR Compilation Pipeline

High Performance Computing (HPC) and Quantum Computing (QC) systems are increasingly converging towards unified High Performance Computing-Quantum Computing (HPCQC) infrastructures, driven by a growing need to bridge classical and quantum workflows, which affects all levels of the system stack, from the hardware to compilers and runtimes, all the way to applications. However, today's QC devices are still in the Noisy Intermediate-Scale Quantum (NISQ) era, are error-prone and resource-limited, and therefore require specialized optimizations and topology mappings to achieve sufficient fidelity. This places special emphasis on proper compilation and optimization within the overall quantum software stack. Authors: Andre Youssefi, Ercüment Kaya, Minh Chung.

Why it matters

Read this for the paper's specific claim in Computer Systems: High Performance Computing (HPC) and Quantum Computing (QC) systems are increasingly converging towards unified High Performance Computing-Quantum Computing (HPCQC) infrastructures, driven by a growing need to bridge classical and quantum workflows, which affects all levels of the system stack, from the hardware to compilers and runtimes, all the way to applications.

Primary paperarxivquant-ph
Original
Computer Systems
arXiv9/24/2026
Recently

Compiler and Hardware Co-Design for Accelerator Architectures

Heterogeneous accelerator architectures offer an efficient path to performance for compute-intensive workloads. However, full-stack integration remains difficult. We present EAAC (Extensible Accelerator Architecture), a flexible and extensible compiler and hardware architecture designed to lower the overhead of hardware-compiler co-development for rapid prototyping of hardware accelerators. Authors: Karl Herman Krause, Emad Jacob Maroun, Martin Schoeberl.

Why it matters

Read this for the paper's specific claim in Computer Systems: Heterogeneous accelerator architectures offer an efficient path to performance for compute-intensive workloads.

Primary paperarxivcs.AR
Original
Computer Systems
GitHub Engineering9/25/2026
Recently

Improving site performance by shipping more CSS

com away from CSS-in-JS. The post Improving site performance by shipping more CSS appeared first on The GitHub Blog.

Why it matters

Read this for the official technical update in Computer Systems: The post Improving site performance by shipping more CSS appeared first on The GitHub Blog.

Official sourceArchitecture & optimizationEngineering
Original
Computer Systems
arXiv9/23/2026
Recently

Flamingo: On Load Balancing in DAG-based Consensus Protocols

Distributed data management systems deployed in untrusted environments rely on Byzantine Fault-Tolerant (BFT) consensus protocols to tolerate malicious failures. DAG-based BFT protocols improve throughput by letting validators disseminate transactions concurrently and by scaling execution across multiple workers. However, imbalances in workload or resource capacity can still degrade performance significantly. Authors: Zhen Ping Khor, Garvit Gupta, Mohammad Javad Amiri.

Why it matters

Read this for the paper's specific claim in Computer Systems: Distributed data management systems deployed in untrusted environments rely on Byzantine Fault-Tolerant (BFT) consensus protocols to tolerate malicious failures.

Primary paperarxivcs.DC
Original
Computer Systems
Microsoft Azure Blog9/23/2026
Recently

Your architecture diagram is not your resilience

For years, resilience was something you set up once. That kept the lights on, but it treated resilience as a project with an end date rather than a property you maintain. The post Your architecture diagram is not your resilience appeared first on Microsoft Azure Blog.

Why it matters

Read this for the official technical update in Computer Systems: For years, resilience was something you set up once.

Official sourceAnalyticsDevOps
Original

Archived section

Developer Tools / Open Source · 5

Developer Tools / Open Source
arXiv9/24/2026
Recently

NEUROTESTGEN: Neuro-Symbolic Guided Test Generation with Large Language Models

Ensuring high structural coverage remains a fundamental challenge in automated test generation, particularly for complex software systems where reaching specific lines or branches requires satisfying intricate control- and data-flow constraints. Large Language Models (LLMs) have recently demonstrated strong capabilities in producing human-like test cases; however, they often struggle to generate inputs that satisfy precise path conditions. Conversely, symbolic execution can systematically derive such constraints, but it often fails to construct realistic, executable test cases and is constrained by scalability limitations. Authors: Ruixin Zhang, Jiho Shin, Hung Viet Pham.

Why it matters

Read this for the paper's specific claim in Developer Tools / Open Source: Ensuring high structural coverage remains a fundamental challenge in automated test generation, particularly for complex software systems where reaching specific lines or branches requires satisfying intricate control- and data-flow constraints.

Primary paperarxivcs.SE
Original
Developer Tools / Open Source
Vercel Blog9/24/2026
Recently

Vercel Sandbox now supports memory observability

Vercel Sandbox observability You can access sandbox memory usage data in the dashboard and through the CLI via the command. vercel metrics The Memory Usage card reports average, P75, and P95 memory across your sandboxes, alongside the existing CPU usage and data transfer metrics on the sandbox overview page and the project and team-level pages. On the sandbox detail page, charts are designed to show how close a sandbox is running to its memory ceiling by: Sandbox memory is also available in the Observability query builder as the measure on the sandbox usage event.

Why it matters

Read this for the official technical update in Developer Tools / Open Source: Vercel Sandbox observability You can access sandbox memory usage data in the dashboard and through the CLI via the command.

Official source
Original
Developer Tools / Open Source
GitHub Blog9/24/2026
Recently

AI-powered fuzzing with the GitHub Security Lab Taskflow Agent

In this blog post, I explain how to use the new fuzzing taskflow based on the GitHub Security Lab Taskflow Agent AI framework. The post AI-powered fuzzing with the GitHub Security Lab Taskflow Agent appeared first on The GitHub Blog.

Why it matters

Read this for the official technical update in Developer Tools / Open Source: In this blog post, I explain how to use the new fuzzing taskflow based on the GitHub Security Lab Taskflow Agent AI framework.

Official sourceAI & MLApplication security
Original
Developer Tools / Open Source
CNCF Blog9/25/2026
Recently

Security Slam 2026 – Fall edition

Security Slam 2026 – Fall Edition is a 30-day virtual event from October 5 through November 6, 2026. What Is the Security Slam? The Open Source Security Foundation (OpenSSF) is partnering with the Cloud Native Computing...

Why it matters

Read this for the official technical update in Developer Tools / Open Source: Security Slam 2026 – Fall Edition is a 30-day virtual event from October 5 through November 6, 2026.

Official sourceBlog
Original
Developer Tools / Open Source
Docker Blog9/24/2026
Recently

Docker and CNCF partner on an open spec for agent permissions

Docker is bringing the open source Sandbox Kit Spec to the CNCF, so AI agent permissions become a neutral, vendor independent standard built on OCI.

Why it matters

Read this for the official technical update in Developer Tools / Open Source: Docker is bringing the open source Sandbox Kit Spec to the CNCF, so AI agent permissions become a neutral, vendor independent standard built on OCI.

Official sourceCompanyPartnerships
Original

Archived section

Cloud / Infrastructure · 5

Cloud / Infrastructure
Cloudflare Blog9/24/2026
Recently

How Cloudflare addressed a cross-tenant data exposure vulnerability in Containers

External security researchers at Accomplish identified a vulnerability in Cloudflare Containers that could expose residual disk data from previous workloads. We explain how the issue worked, how we investigated it, and the steps we took to remediate it.

Why it matters

Read this for the official technical update in Cloud / Infrastructure: External security researchers at Accomplish identified a vulnerability in Cloudflare Containers that could expose residual disk data from previous workloads.

Official sourceContainersDeveloper Platform
Original
Cloud / Infrastructure
IEEE Spectrum9/24/2026
Recently

Engineering the Substation Exit for Reliability, Capacity, and Expansion

This white paper gives distribution engineers and utility teams a practical overview of the substation exit, the short and high-consequence section where station capacity divides into individual feeders. It explains how covered, spacer-supported overhead construction can reduce common contact-driven faults while leaving room for future circuits. What you will learn about: Why a fault near the substation exposes more customers than one farther along the feeder, and why the first spans out of the station carry so much reliability weight.

Why it matters

Read this for the concrete reporting in Cloud / Infrastructure: This white paper gives distribution engineers and utility teams a practical overview of the substation exit, the short and high-consequence section where station capacity divides into individual feeders.

Trusted sourceType-whitepaperCables
Original
Cloud / Infrastructure
IEEE Spectrum9/24/2026
Recently

Mexican EPICS in IEEE Team Builds Portable Educational Platform

In Guadalajara, Mexico, many high schools have motivated teachers and talented students with an interest in science, technology, engineering, and mathematics, but they lack access to advanced tools such as robotics laboratories. The resources shortfall limits the students' opportunities for hands-on learning on cutting-edge applications. A team from ITESO, Universidad Jesuita de Guadalajara, is working to change that.

Why it matters

Read this for the concrete reporting in Cloud / Infrastructure: In Guadalajara, Mexico, many high schools have motivated teachers and talented students with an interest in science, technology, engineering, and mathematics, but they lack access to advanced tools such as robotics laboratories.

Trusted sourceRoboticsAi
Original
Cloud / Infrastructure
Vercel Blog9/24/2026
Recently

The Vercel Bug Bounty Program is now publicly available

In 2022, we launched a private bug bounty program through HackerOne. Over the last several years, we've worked with HackerOne's VIP program to refine our process, targets, and scope, onboarding thousands of their best researchers and hardening security across our platform. We've translated those learnings into several public bounty programs: Today, we are combining our private and OSS bounty programs into a single, public.

Why it matters

Read this for the official technical update in Cloud / Infrastructure: In 2022, we launched a private bug bounty program through HackerOne.

Official source
Original
Cloud / Infrastructure
Docker Blog9/24/2026
Recently

From Dockerfile to Kit: the Docker Sandboxes Kit Specification

Docker's Sandbox Kit Specification v3 packages an AI agent's network rules, credentials, and volumes as an ordinary, pinnable OCI image.

Why it matters

Read this for the official technical update in Cloud / Infrastructure: Docker's Sandbox Kit Specification v3 packages an AI agent's network rules, credentials, and volumes as an ordinary, pinnable OCI image.

Official sourceEngineeringProducts
Original