Intelligence Digest

OpenAI Blog · 2026-04-02

OpenAI acquires TBPN to accelerate global conversations around AI and support independent media, expanding dialogue with builders, businesses, and the broader tech community.

OpenAI Blog · 2026-04-02

Codex now includes pay-as-you-go pricing for ChatGPT Business and Enterprise, providing teams a more flexible option to start and scale adoption.

OpenAI Blog · 2026-04-01

Gradient Labs uses GPT-4.1 and GPT-5.4 mini and nano to power AI agents that automate banking support workflows with low latency and high reliability.

OpenAI Blog · 2026-03-31

OpenAI raises $122 billion in new funding to expand frontier AI globally, invest in next-generation compute, and meet growing demand for ChatGPT, Codex, and enterprise AI.

Hacker News · 2026-04-02

Comments

Hacker News · 2026-04-02

Comments

Hacker News · 2026-04-02

Comments

Hacker News · 2026-04-02

Comments

GitHub Blog · 2026-04-01

Recent attacks on open source focus on exfiltrating secrets; here are the prevention steps you can take today, plus a look at the security capabilities GitHub is working on. The post Securing the open source supply chain across GitHub appeared first on The GitHub Blog .

GitHub Blog · 2026-04-01

/fleet lets Copilot CLI dispatch multiple agents in parallel. Learn how to write prompts that split work across files, declare dependencies, and avoid common pitfalls. The post Run multiple agents at once with /fleet in Copilot CLI appeared first on The GitHub Blog .

GitHub Blog · 2026-03-31

I used coding agents to build agents that automated part of my job. Here's what I learned about working better with coding agents. The post Agent-driven development in Copilot Applied Science appeared first on The GitHub Blog .

ScienceDaily — Artificial Intelligence · 2026-03-31

DNA robots are emerging as tiny programmable machines that could one day deliver drugs, hunt viruses, and build molecular-scale devices. By borrowing ideas from traditional robotics and combining them with DNA folding techniques, scientists are creating structures that can move and act with precision. These robots can be guided using chemical reactions or external signals like light and magnetic fields.

arXiv · 2026-04-01

We introduce SoftAct, a framework for teaching soft robot hands to perform human-like manipulation skills by explicitly reasoning about contact forces. Leveraging immersive virtual reality, our system captures rich human demonstrations, including hand kinematics, object motion, dense contact patches, and detailed contact force information. Unlike conventional approaches that retarget human joint trajectories, SoftAct employs a two-stage, force-aware retargeting algorithm. The first stage attributes demonstrated contact forces to individual human fingers and allocates robot fingers proportionally, establishing a force-balanced mapping between human and robot hands. The second stage performs online retargeting by combining baseline end-effector pose tracking with geodesic-weighted contact refinements, using contact geometry and force magnitude to adjust robot fingertip targets in real time

arXiv · 2026-04-01

We present HippoCamp, a new benchmark designed to evaluate agents' capabilities on multimodal file management. Unlike existing agent benchmarks that focus on tasks like web interaction, tool use, or software automation in generic settings, HippoCamp evaluates agents in user-centric environments to model individual user profiles and search massive personal files for context-aware reasoning. Our benchmark instantiates device-scale file systems over real-world profiles spanning diverse modalities, comprising 42.4 GB of data across over 2K real-world files. Building upon the raw files, we construct 581 QA pairs to assess agents' capabilities in search, evidence perception, and multi-step reasoning. To facilitate fine-grained analysis, we provide 46.1K densely annotated structured trajectories for step-wise failure diagnosis. We evaluate a wide range of state-of-the-art multimodal large langu

arXiv · 2026-04-01

The rise of test-time scaling has remarkably boosted the reasoning and agentic proficiency of Large Language Models (LLMs). Yet, standard Transformers struggle to scale inference-time compute efficiently, as conventional looping strategies suffer from high computational overhead and a KV cache that inflates alongside model depth. We present Universal YOCO (YOCO-U), which combines the YOCO decoder-decoder architecture with recursive computation to achieve a synergistic effect greater than either alone. Built on the YOCO framework, YOCO-U implements a Universal Self-Decoder that performs multiple iterations via parameter sharing, while confining the iterative process to shallow, efficient-attention layers. This combination yields a favorable capability-efficiency tradeoff that neither YOCO nor recursion achieves independently. The YOCO architecture provides a constant global KV cache and l

arXiv · 2026-04-01

The thermodynamic resourcefulness of quantum channels primarily depends on their underlying causal structure and their ability to generate quantum correlations. We quantify this interplay within the resource theory of athermality for bipartite quantum channels in the presence of a side channel acting as memory, referred to as the resource theory of conditional athermality. For channels with trivial output Hamiltonians, we characterize the optimal one-shot rates for distilling the identity gate from a given channel, as well as the cost of simulating the channel using the identity gate, under conditional Gibbs-preserving superchannels. We show that these rates have a direct trade-off relation with the conditional channel entropies, attributing operational significance to signaling in quantum processes. Furthermore, we establish an equipartition property for the conditional channel min-entr

arXiv · 2026-04-01

Reconstructing full spatio-temporal dynamics from sparse observations in both space and time remains a central challenge in complex systems, as measurements can be spatially incomplete and can be also limited to narrow temporal windows. Yet approximating the complete spatio-temporal trajectory is essential for mechanistic insight and understanding, model calibration, and operational decision-making. We introduce LAPIS-SHRED (LAtent Phase Inference from Short time sequence using SHallow REcurrent Decoders), a modular architecture that reconstructs and/or forecasts complete spatiotemporal dynamics from sparse sensor observations confined to short temporal windows. LAPIS-SHRED operates through a three-stage pipeline: (i) a SHRED model is pre-trained entirely on simulation data to map sensor time-histories into a structured latent space, (ii) a temporal sequence model, trained on simulation-

arXiv · 2026-04-01

AI weather prediction has advanced rapidly, yet no unified mathematical framework explains what determines forecast skill. Existing theory addresses specific architectural choices rather than the learning pipeline as a whole, while operational evidence from 2023-2026 demonstrates that training methodology, loss function design, and data diversity matter at least as much as architecture selection. This paper makes two interleaved contributions. Theoretically, we construct a framework rooted in approximation theory on the sphere, dynamical systems theory, information theory, and statistical learning theory that treats the complete learning pipeline (architecture, loss function, training strategy, data distribution) rather than architecture alone. We establish a Learning Pipeline Error Decomposition showing that estimation error (loss- and data-dependent) dominates approximation error (arch

arXiv · 2026-04-01

Efficient robotic extraterrestrial exploration requires robots with diverse capabilities, ranging from scientific measurement tools to advanced locomotion. A robotic team enables the distribution of tasks over multiple specialized subsystems, each providing specific expertise to complete the mission. The central challenge lies in efficiently coordinating the team to maximize utilization and the extraction of scientific value. Classical planning algorithms scale poorly with problem size, leading to long planning cycles and high inference costs due to the combinatorial growth of possible robot-target allocations and possible trajectories. Learning-based methods are a viable alternative that move the scaling concern from runtime to training time, setting a critical step towards achieving real-time planning. In this work, we present a collaborative planning strategy based on Multi-Agent Prox

arXiv · 2026-04-01

As LLM agents tackle increasingly complex tasks, a critical question is whether they can maintain strategic coherence over long horizons: planning under uncertainty, learning from delayed feedback, and adapting when early mistakes compound. We introduce $\texttt{YC-Bench}$, a benchmark that evaluates these capabilities by tasking an agent with running a simulated startup over a one-year horizon spanning hundreds of turns. The agent must manage employees, select task contracts, and maintain profitability in a partially observable environment where adversarial clients and growing payroll create compounding consequences for poor decisions. We evaluate 12 models, both proprietary and open source, across 3 seeds each. Only three models consistently surpass the starting capital of \$200K, with Claude Opus 4.6 achieving the highest average final funds at \$1.27 M, followed by GLM-5 at \$1.21 M

arXiv · 2026-04-01

We study the problem of controlling how a limited communication bandwidth budget is allocated across heterogeneously quantized sensor measurements. The performance criterion is the trace of the error covariance matrix of the linear minimum mean square error (LMMSE) state estimator, i.e., an $A$-optimal design criterion. Minimizing this criterion with a bit budget constraint yields a nonconvex optimization problem. We derive a formula that reduces each evaluation of the gradient to a single Cholesky factorization. This enables efficient optimization by both a projection-free Frank-Wolfe method (with a computable convergence certificate) and an interior point method with L-BFGS Hessian approximation over the problem's continuous relaxation. A largest remainder rounding procedure recovers integer bit allocations with a bound on the quality of the rounded solution. Numerical experiments in I

arXiv · 2026-04-01

Scientific algorithm discovery is iterative: hypotheses are proposed, implemented, stress-tested, and revised. Current LLM-guided search systems accelerate proposal generation, but often under-represent scientific structure by optimizing code-only artifacts with weak correctness/originality gating. We present CliffSearch, an agentic evolutionary framework in which the core evolution operators (pair selection, crossover, mutation, and review) are implemented as LLM agents, and the loop is designed around three principles: (1) each node is a structured scientific artifact, instantiated in either theory+code or code_only mode, (2) reviewer judgments of correctness and originality are first-class selection gates alongside optimization of the benchmark metric of interest, and (3) mutation is split into exploration and correction pathways with distinct objectives. Exploration mutation imports

NVD · 2024-03-08

The issue was addressed with improved memory handling. This issue is fixed in iOS 17.4 and iPadOS 17.4, macOS Sonoma 14.4, tvOS 17.4, visionOS 1.1, watchOS 10.4. Processing web content may lead to arbitrary code execution.

NVD · 2024-01-23

A type confusion issue was addressed with improved checks. This issue is fixed in Safari 17.3, iOS 15.8.7 and iPadOS 15.8.7, iOS 16.7.5 and iPadOS 16.7.5, iOS 17.3 and iPadOS 17.3, macOS Monterey 12.7.3, macOS Sonoma 14.3, macOS Ventura 13.6.4, tvOS 17.3, visionOS 1.0.2. Processing maliciously crafted web content may lead to arbitrary code execution. This fix associated with the Coruna exploit was shipped in iOS 17.3 on January 22, 2024. This update brings that fix to devices that cannot update to the latest iOS version.

NVD · 2024-01-23

The issue was addressed with improved memory handling. This issue is fixed in Safari 17.3, iOS 16.7.5 and iPadOS 16.7.5, iOS 17.3 and iPadOS 17.3, macOS Sonoma 14.3, tvOS 17.3, watchOS 10.3. Processing web content may lead to arbitrary code execution.

NVD · 2024-01-23

Multiple memory corruption issues were addressed with improved memory handling. This issue is fixed in iOS 16.7.5 and iPadOS 16.7.5, iOS 17.3 and iPadOS 17.3, macOS Sonoma 14.3. Processing maliciously crafted web content may lead to arbitrary code execution.

NVD · 2024-01-23

The issue was addressed with improved memory handling. This issue is fixed in macOS Sonoma 14.3. Processing web content may lead to arbitrary code execution.

NVD · 2024-05-14

The issue was addressed with improved checks. This issue is fixed in macOS Sonoma 14.5. An app may be able to execute arbitrary code out of its sandbox or with certain elevated privileges.

NVD · 2024-06-11

The issue was addressed with improved checks. This issue is fixed in macOS Monterey 12.7.4, macOS Sonoma 14.4, macOS Ventura 13.6.5. An app may be able to break out of its sandbox.

NVD · 2024-03-08

The issue was addressed with improved checks. This issue is fixed in iOS 16.7.6 and iPadOS 16.7.6, iOS 17.4 and iPadOS 17.4, macOS Sonoma 14.4, macOS Ventura 13.6.5, tvOS 17.4, watchOS 10.4. An app may be able to break out of its sandbox.

NVD · 2024-03-08

This issue was addressed by removing the vulnerable code. This issue is fixed in iOS 16.7.6 and iPadOS 16.7.6, iOS 17.4 and iPadOS 17.4, macOS Sonoma 14.4, tvOS 17.4, visionOS 1.1, watchOS 10.4. An app may be able to break out of its sandbox.

NVD · 2024-03-08

The issue was addressed with improved memory handling. This issue is fixed in iOS 17.4 and iPadOS 17.4, macOS Sonoma 14.4, tvOS 17.4, watchOS 10.4. An app may be able to execute arbitrary code out of its sandbox or with certain elevated privileges.

NVD · 2024-05-14

The issue was addressed with improved checks. This issue is fixed in macOS Sonoma 14.5. An app may be able to execute arbitrary code with kernel privileges.

NVD · 2024-05-14

A logic issue was addressed with improved checks. This issue is fixed in macOS Monterey 12.7.5, macOS Sonoma 14.5, macOS Ventura 13.6.7. An app may be able to elevate privileges.

NVD · 2024-05-14

The issue was addressed with improved memory handling. This issue is fixed in macOS Sonoma 14.5. Processing a file may lead to unexpected app termination or arbitrary code execution.

NVD · 2024-05-14

A logic issue was addressed with improved restrictions. This issue is fixed in macOS Sonoma 14.5. An app may be able to gain root privileges.

Brave Search

2. Define the LoRA configuration as discussed above. 3. Define the train and test splits of the prepped instruction following data into Hugging Face Dataset objects. 4. Define training arguments. These include the number of epochs, batch size and other training hyperparameters which will be kept constant during this exercise. 5. Pass these arguments into an instance of SFTTrainer. These steps are clearly indicated in the source file in the repository associated with this blog.

Brave Search

<strong>This repository contains code for fine-tuning permissive open source LLMs using low-rank adaptation (LoRA).</strong> Code is tested using Stanford Alpaca dataset. Estimated training time for fine-tuning RedPajama-INCITE-Base-7B-v0.1 with a single RTX ...

Brave Search

Repository of course materials for a multi-day course on machine learning for tabular data using Scikit-Learn and XGBoost - davidrpugh/machine-learning-for-tabular-data

Brave Search

We want to create a practical guide to developing quality predictive models from tabular data. We’ll publish materials here as we create them and welcome community contributions in the form of discussions, suggestions, and edits. We also want these materials to be reusable and open. The sources are in the source GitHub repository with a Creative Commons license attached (see below).

Brave Search

Read more about Salesforce Marketing Cloud Account Engagement ... Open Source Free tool. Enjoy all the features of Clarity at absolutely zero cost. You&#x27;ll never run into traffic limits or be forced to upgrade to a paid version. ... Open Source Per number of page views. ... Open Source Contact Bloomreach for pricing information. ... Maximize your conversion with riyo.ai, an all-in-one tool.

Brave Search

Open online course on JavaScript based modern web development by University of Helsinki and Houston Inc..

Brave Search

Use open source VM detonation services like hybrid-analysis.com or malwr.com to get the behavior quickly · Use the VM detonation service to capture any network connections or packet data. If you can&#x27;t do this then we will need to dynamically debug the malware.

Brave Search

Malware Analysis on a Budget - Discussion of malware analysis tools and research projects out in the open-source community · MalwareDNA - Talk about an instruction-analysis technique I devised in 2013 · Scott Nusbaum taught this class in Spring 2019, here is his great curriculum · Adventures in Security (http://securitykitten.github.io/) - Nick Hoffman: A colleage, former coworker, and friend · RPISEC’s Malware Analysis Course - Rensselaer Polytechnic Institute

Brave Search

<strong>The full-scale pre-training code (providing conversion, construction, and loading of large corpora) and LoRA instruction fine-tuning code are open-sourced (support multi-machine multi-GPU).</strong>

Brave Search

Linux Foundation’s Fundamentals of Quantum Computing course. These courses blend theoretical concepts with hands-on programming, mostly using open-source tools, helping learners gain practical skills to run quantum algorithms on simulators and real devices.