Knowledge catalogue
arXiv:2605.21706v2 Announce Type: replace Abstract: Safety-aligned language models are trained to refuse harmful requests, yet refusal behavior can be suppressed by steering their internal representat
arXiv:2408.15344v2 Announce Type: replace Abstract: Many scientific and engineering problems involve observing a common phenomenon through multiple heterogeneous sensors or measurement modalities. Suc
arXiv:2606.01374v1 Announce Type: new Abstract: Observable performance is commonly used to characterize biological systems. In adaptive systems, however, similar performances may arise from distinct o
arXiv:2606.02486v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models generalize across static manipulation but fail when objects move during task execution. They map the current observa
NVIDIA announced three new offerings: Cosmos 3, an advanced video generation model; Nemotron 3 Ultra, an upgraded language model; and RTX Spark, likely a tool or framework for developers. These releas
GitHub is developing AI agents to automate software development workflows and enhance developer productivity. The discussion likely covers GitHub's strategy for integrating autonomous AI capabilities
arXiv:2607.26460v1 Announce Type: new Abstract: Mobile manipulation requires generating whole-body action chunks that jointly satisfy goal reaching, collision avoidance, base kinematic constraints, ma
arXiv:2607.23837v1 Announce Type: cross Abstract: Large language models generalize well to individual tasks but lack an inherent mechanism for learning them sequentially, leading to catastrophic forge
arXiv:2607.24569v1 Announce Type: cross Abstract: Model-based active flow control requires predictive models that are accurate, stable, and fast enough for real-time optimisation. In controlled wake f
OpenAI’s Codex, an AI‑powered coding engine, has surged beyond the developer community to reach over 10 million users—combining Codex with its new “ChatGPT Work” platform that turns the model into a g
arXiv:2606.04249v1 Announce Type: new Abstract: Prospective reconstruction is crucial in many clinical applications such as MRI-guided radiotherapy, which demands accurate image reconstruction and fas
This AI News summary covers recent developments in image generation models, specifically discussing Reve 2 and Ideogram 4's layout capabilities for creating structured visual designs. The piece likely
Lukas Petersson and Axel Backlund of Andon Labs discuss their work on evaluation methods and benchmarking for AI systems, likely covering approaches to assessing AI capabilities and safety in real-wor
arXiv:2605.27852v1 Announce Type: cross Abstract: Unified and scalable Transformers have recently achieved remarkable success in modeling diverse phenomena traditionally associated with computer graph
arXiv:2605.17284v1 Announce Type: cross Abstract: End-to-end autonomous driving systems powered by Vision-Language-Action (VLA) models achieve strong performance on common driving scenarios, yet remai
arXiv:2510.04006v2 Announce Type: replace Abstract: Data-driven machine learning (ML) models are reshaping weather forecasting and have shown the potential to accelerate and surpass traditional physic
arXiv:2605.15862v1 Announce Type: new Abstract: Adaptive biomechanical systems may show similar observable gait performance while differing in latent organization and longitudinal behavior. This study
arXiv:2608.08029v1 Announce Type: cross Abstract: Khatri et al. (2026) [DOI: 10.1109/DSN-W70714.2026.00027] show that lightweight MLP probes on final-layer activations of a single 8B model (LLaMA-3.1-
arXiv:2606.31268v1 Announce Type: cross Abstract: The growing demand for privacy-preserving data sharing has positioned synthetic data generation as a critical component of responsible AI workflows. D
This dispatch covers developments from the AI Engineer World Fair, likely discussing emerging trends in AI engineering including feedback loops in AI systems, the concept of software factories powered
This Latent Space article discusses recent AI model releases, covering Anthropic's Claude Sonnet 5 and anticipated details about Fable 5, likely examining their capabilities, improvements, and implica
Autoresearch describes a feedback mechanism where AI agents can evaluate their own outputs and use that introspection to iteratively improve their reasoning and decision-making capabilities. This conc
Forward-deployed engineers are specialized software professionals embedded within client organizations to provide hands-on technical expertise and implementation support for AI and software solutions.
Cursor, an AI-powered code editor, implements enterprise deployment strategies to integrate AI-assisted development tools into corporate environments. The discussion likely covers Cursor's architectur
This episode discusses cutting-edge diffusion model research applications beyond large language models, featuring insights from Evan Feinberg and Sergey Edunov of Genesis Molecular AI on how diffusion
Zach Lloyd, CEO of Warp, discusses the evolution of software development toward 'software factories'—automated systems that leverage AI to streamline coding workflows and increase developer productivi
arXiv:2602.15382v2 Announce Type: replace Abstract: Multi-Agent Systems (MAS) powered by Large Language Models have unlocked advanced collaborative reasoning, yet they remain bottlenecked by discrete
arXiv:2605.18918v1 Announce Type: cross Abstract: Modern AI assistants are agentic. To answer a single user request, the underlying language model pulls in information from many sources, such as web s
arXiv:2605.06225v2 Announce Type: replace-cross Abstract: Steering large language models (LLMs) is usually done by either instruction prompting or activation steering. Prompting often gives strong con
arXiv:2602.09034v2 Announce Type: replace-cross Abstract: Neuroimaging does not observe causal variables directly: hemodynamics and volume conduction distort signals so that statistical dependence nee
arXiv:2604.08636v1 Announce Type: cross Abstract: Designing robot morphologies and kinematics has traditionally relied on human intuition, with little systematic foundation. Motion-design co-optimizat
ChatGPT Work was launched by OpenAI on July 9, 2026 as an agent‑oriented knowledge‑work platform that combines chat, Codex tools and cloud agents across fourteen model configurations. Within three wee
Poolside AI’s co‑CEO Eiso Kant explains that a small, top‑research team built a model‑factory capable of training Laguna S, a 118‑billion‑parameter mixture‑of‑experts model that outperforms the ~1 tri
According to this Latent Space article, OpenAI has launched GPT 5.6 with variants named Sol, Terra, and Luna, while integrating Codex capabilities into ChatGPT to create a broader superapp platform. T
SpaceX's AI division has launched Grok 4.5, positioned as their first Opus-class model following the acquisition of Cursor. The release represents a significant capability upgrade in SpaceX's AI offer
Lilian Weng provides a comprehensive summary of 35 research papers focusing on Harness Engineering approaches for Reasoning and System Integration (RSI), covering recent developments in this specializ
This article discusses how AI infrastructure needs to adapt to support autonomous agents effectively, with insights from Modal's CTO on the technical and architectural requirements for agent deploymen
This article from Latent Space provides an overview of Fable, covering its capabilities, architecture, or use cases as an AI model or framework. The guide likely serves as a reference resource for dev
This dispatch from Latent Space discusses debates within the AI engineering community regarding 'loops' in AI systems and the current state of AI engineering practices. The article likely covers techn
Andrew Qu from Vercel discusses how AI agents represent a fundamentally different category of software compared to traditional applications, exploring their unique characteristics and implications for
This dispatch examines the emerging capabilities and challenges of AI autoresearch systems, analyzing the tension between autonomous AI agents conducting research and the preservation of human agency
This article discusses 'skill engineering' as an approach to AI system design that emphasizes iterative development and refinement of AI capabilities, arguing against the 'one-shot' design philosophy
This article discusses how future websites could dynamically generate personalized content and layouts for each individual visitor using AI and generative technologies, rather than serving static page
Ahmad Osman discusses the technological and practical reasons why locally-run AI models are becoming increasingly competitive with cloud-based alternatives, likely covering improvements in model effic
OpenAI has released GPT-5.6 with three variants (Sol, Terra, Luna) in a restricted beta program limited to trusted partners, according to reporting from Latent Space. The restricted access model sugge
OpenAI reported significant increases in internal Codex token usage across different departments between November 2025 and the reporting period, with Research experiencing the largest growth at 56x, f
Meta released Harness, an open-source framework for evaluating and benchmarking AI model performance across diverse tasks and datasets. The tool enables standardized testing of language models and aim
Anthropic's Claude API now supports multiplayer agent capabilities that enable multiple Claude instances to collaborate within Slack, with features for proactive behavior and persistent state manageme
This article discusses why Databricks leaders argue that the frontier AI ecosystem should remain open rather than closed, likely covering topics such as open-source AI development, the benefits of col
This article analyzes SpaceX's valuation and business model through the lens of AI infrastructure, arguing that SpaceX functions as a 'Neocloud' (a next-generation cloud computing platform) generating
This episode likely discusses red-teaming methodologies and adversarial testing approaches in AI systems, featuring security researchers Zico Kolter and Matt Fredrikson from Carnegie Mellon University
Latent Space is offering an exclusive promotion providing $250 off tickets to an AI Engineer event, with the discount valid until Monday. This appears to be a time-limited offer announced through thei
This episode discusses the growing landscape of open-source AI models, contrasts the approaches and philosophies of Model Labs versus Agent Labs in AI development, and explores fundamental limitations
This article from Latent Space discusses Anthropic's Claude Fable 5 model, examining its capabilities in handling creative and mythological content while maintaining safety guardrails, along with cove
FrontierCode is a benchmarking framework designed to evaluate code quality and move beyond 'slop'—low-quality or poorly generated code—in AI-generated outputs. The framework likely provides metrics an
This article discusses best practices for ensuring high-quality reinforcement learning (RL) environments, addressing common pitfalls that lead to poor simulation design and implementation. It likely p
Microsoft announced the MAI-Thinking-1 and MAI family of models at their Build conference, representing new advances in their AI model lineup. These models likely focus on improved reasoning capabilit
This episode features Satya Nadella, CEO of Microsoft, in a crossover special between the 'No Priors' and 'Latent Space' podcasts, likely discussing Microsoft's AI strategy, product developments, and
Carina Hong discusses how Axiom Math is scaling beyond informal AI approaches to create more rigorous, systematic methods for AI-assisted mathematics education and problem-solving. The episode likely
Video agent models represent the next frontier in AI by extending language model capabilities to process, understand, and act on video content in real-time, enabling autonomous agents to perceive and