AI Daily Briefing

The following article originally appeared on Addy Osmani’s blog and is being republished here with the author’s permission. A software factory harnesses loops at scale. You can run the loop with humans in it (light factory), trading judgment and concentration against speed and breakage. Or you can ignore the humans (dark factory) and let those […]

Migrate a multi-model healthcare AI agent from self-managed Amazon ECS with AWS Fargate to Amazon Bedrock AgentCore runtime, preserving triple-model orchestration and vector-enhanced knowledge retrieval while reducing infrastructure management. The framework-agnostic pattern applies across healthcare, financial services, and manufacturing.

Today we are announcing the new AgentCore runtime, a capability of Amazon Bedrock AgentCore built for the speed, flexibility, and cost efficiency that production agents demand. It reclaims memory as sessions release it and delivers consistent cold starts regardless of image size or concurrency.

Deploy production-ready Hugging Face models on Amazon SageMaker AI using six open-source agent skills. Point a coding agent at a model and get back a real-time endpoint with the right serving container, autoscaling, Amazon CloudWatch alarms, and a verified teardown path.

We dive into these questions and other AI hot takes on the latest episode of the GitHub Podcast. The post Should you read the code, is RAG dead, and did Skills kill MCP? appeared first on The GitHub Blog.

Google worked side-by-side with designers Jane Wade and Sergio Hudson to custom-design Google Flow tools to prep for NYFW.

Sponsored Content It's no secret that AI agents burn massive amounts of tokens on search results and file retrievals. They pull in...

Which open-source harness works with Ollama, LM Studio, or llama.cpp? 11 verified picks with licenses and setup rules. The post Best Open-Source Agent Harnesses for Local LLMs in 2026 appeared first on MarkTechPost.

Git made isolated development a baseline for software teams. Each developer can create...

Alibaba's Qwen3.8-Omni-Flash understands audio and video, plans tasks, calls tools, and reports about 45.7% fewer tokens on OmniVideoBench. The post Alibaba Qwen Releases Qwen3.8-Omni-Flash: A 1M-Context Omni-Modal Model Built Around Agentic Audio-Video Understanding and Tool Use appeared first on MarkTechPost.

Building an AI prototype is easy, but operating autonomous agents at scale requires production-grade tooling. Salesforce Agentforce bridges the gap from "vibe coding" to enterprise reliability by combining synthetic stress-testing, real-time optimization, dynamic agentic UIs, and deterministic guardrails—as proven by Southwest Airlines' 7x ROI. The post Salesforce Agentforce: Bridging the Enterprise AI Gap from ‘Vibe Coding’ to Battle-Tested Orchestration appeared first on MarkTechPost.

Discussion | Link

a quiet day

Discussion | Link

arXiv:2609.19377v1 Announce Type: new Abstract: Recovering numerical series from line plots requires accurate axis calibration and reliable curve extraction. We present LinePilot Digitizer (LinePilot), which combines continuous color-based curve recovery with three calibration modes: LinePilot (standard), LinePilot (enhanced), and LinePilot (OCR). We also introduce DigitizerBench, the first dedicated benchmark for systematically evaluating digitizer performance, using an orthogonal design spanning signal, rendering, and plot-structure factors with complementary automatic and human-guided evaluations. We evaluate performance using failure-penalized capped normalized root-mean-square error (FPC-NRMSE), which assigns unit loss to missing, unusable, or catastrophically inaccurate outputs. On…

arXiv:2609.19359v1 Announce Type: new Abstract: Falls in older adults are often preceded by changes in mobility, balance, and postural transitions. This paper presents a wireless smart insole platform and machine-learning workflow for recognizing sitting, standing, walking, and unstable walking from plantar-pressure and inertial signals. Each insole integrates 16 active pressure-sensing locations and a six-dimensional IMU stream consisting of tri-axial acceleration and angular velocity. Data were collected from 15 healthy adults at 80~Hz and segmented into overlapping windows. Window length and candidate model families were first screened with stratified 10-fold cross-validation; the primary performance estimate was then obtained with participant-independent 5-fold Stratified Group cross-…

arXiv:2609.19425v1 Announce Type: new Abstract: Tool-augmented large language model (LLM) agents fail in a way no tool-selection or tool-security method addresses: they call tools that do not exist and pass arguments no schema declares. Existing defenses either pick the right tool (selection) or constrain what an agent may do with real tools (gating), both of which presuppose the emitted call refers to a real tool at all. We show this is a structural blind spot: a hallucinated call is by construction not a decision any gate made, so no gate can reject it. This paper is primarily a measurement and benchmark study. We give a five-class taxonomy of tool hallucination (H1-H5) and, as a reference point, the Resolution Rung: a training-free, closed-world resolver (registry membership plus a sig…

arXiv:2609.19391v1 Announce Type: new Abstract: LLM coding agents now generate complex programs at a scale that makes thorough human review increasingly difficult, raising the risk of safety and security failures. Common approaches, including fuzz testing, static analysis, and LLM-as-a-Verifier, can detect many failures but struggle to cover all possible edge cases. Formal verification addresses this by providing machine-checkable guarantees over specified properties, but traditionally demands substantial manual specification and proof engineering. We introduce a unified multi-agent framework, MAGS, that generates executable programs with formal safety guarantees, using Dafny as a verification-aware intermediate representation where safety properties can be mechanically checked. MAGS form…

arXiv:2609.19387v1 Announce Type: new Abstract: Agents are increasingly asked to design hardware, and increasingly reported to succeed. Such reports establish that a design improved; they cannot establish why. An agent that improves an accelerator may be reasoning about the machine, or may be searching competently over knobs whose meaning it never recovers -- and only the first transfers to the next architecture. Existing evaluations cannot tell the two apart, because they vary the agent while holding the framing of the problem fixed. We do the opposite. AutoTuring hands the same agent the same 15-dimensional accelerator space twice: once as named architectural knobs with simulator counters, once as anonymous variables on [0,1], with the evaluator, the legal space and the reachable optima…

arXiv:2609.19244v1 Announce Type: new Abstract: Conversational LLM agents increasingly rely on Web search, yet the end-to-end lifecycle of agentic search remains poorly understood. We present the first study of Web search across four major conversational platforms (ChatGPT, Claude, Grok, and DeepSeek), combining real-world user interactions (invivo) with controlled experiments using the same platform's models by their APIs (invitro). We investigate the quality of agentic decisions to invoke Web search, their strategies to formulate queries, the potential domain preferences in the search results they receive, and the choices they make when transforming search results into grounded responses. We find that Web-search decisions vary substantially across platforms and models, while more freque…

Discussion | Link

Discussion | Link

Inside a global bank's shift to self-serve dedicated inference: how Together's DMI gave engineering teams direct control over scaling, models, and testing.

Be alert: targeted attacks on prominent Rustaceans Important warning from Adam Harvey and the crates security team: We believe that there is an ongoing campaign targeting rust-lang members and owners of popular crates that is attempting to compromise devices and accounts in order to use them to publish malware. A video call is set up for something positive — maybe for a job, maybe for a project, maybe for a contract opportunity — and then that's used as a vector to either get the target to install something on their computer (such as a purportedly missing audio codec) or execute another command (for example, via putting a command on the clipboard). Last month this trick was used in a successful supply chain attack against the array ref crate, among others. Any piece of software that depen…

Self-generated prompt injections in compaction summaries In Our framework for reporting model misalignment OpenAI provide "six reports on unexpected or concerning model behavior we’ve observed in the last six months". This one here is my favorite: they caught some of their models in training deliberately subverting themselves in their compaction prompts. Compaction is the process agent systems use when they are running out of tokens in their context window, so they summarize everything that has gone before so they can keep going with more token headroom. In one of the observed instances, a model undergoing reinforcement learning was working on a task to update an existing HTTP API endpoint with a new feature. The model compacted its work so far, and then added the following text to the su…

Anthropic redesigned Projects in Claude Code. The old project was a folder: some files plus one chat. The new one is a single ongoing conversation where Claude acts as coordinator. You describe work, and Claude decides what becomes a thread. Each thread is a full Claude Code cloud session running on its own branch and […] The post Anthropic Launches Claude Code Projects in Beta: Parallel Cloud Sessions That Keep Running After You Close Your Laptop appeared first on MarkTechPost.

Education Innovation

Google and the UN system have launched the UN System Data Commons, a new open platform making global statistics accessible and easy to search.

Discussion | Link

Remember when tech leaders would tell their employees to “move fast and break things”? It seemed that would be the way of AI too. But after a summer where rogue AI agents became reality, and researchers warned that AI could kill us all, a number of leading US AI companies are publicly suggesting it’s time to pump the brakes and “pace the frontier” of bleeding-edge AI development. Their motivations are suspect, but leaders at major AI companies — including Anthropic, OpenAI, Google, Microsoft, and X — are at least paying lip service to the idea of a superintelligence slowdown. Will these AI companies actually slow down? Will anyone step in to regulate these companies like they claim to have wanted for years? Will they manage to convince world leaders that the US must “beat China” to superi…

The revamped projects feature in Claude Code allows users to run multiple agents under the same roof, with a shared memory, goals, and library of files and artifacts. Similar to Grok Bot and other tools that manage groups of AI agents, each project has "threads" running different tasks in parallel, with a "coordinator" directing everything: Under the hood, each thread is a Claude Code cloud session working on its own branch and copy of the repo. The coordinator keeps work organized, but if any threads work on the same code, the overlap is resolved as a merge conflict just like any other PR. Each thread can further split its delegated work … Read the full story at The Verge.

Rust is seemingly the language of the moment, with open-source projects and companies forming an orderly queue to move core The post GitHub and Anthropic used their own agents for major Rust rewrites — with very different playbooks appeared first on The New Stack.

The five criteria for evaluating a database for AI agents are branch isolation, serverless...

Discussion | Link

An agent that needs the outside worldAn engineer at a software company is building...

Deep Life Sci is LangChain's open source agentic assistant for clinical and lab scientists. It pulls from 600K+ ClinicalTrials.gov studies, 29M PubMed abstracts, and 12M PubMed Central full-text articles, with sandboxed sub-agents for real data analysis.

Artificial Intelligence for IT Operations (AIOps) applies AI and machine learning to IT operations to detect anomalies...

Use Case Use Case: Manufacturing Industry Sep 1, 2026 Use Case Use Case: Manufacturing Industry Sep 1, 2026 Authors Unstructured In this article Join our newsletter to receive updates about our features. In this article…

Use Case Use Case: Insurance Industry Sep 1, 2026 Use Case Use Case: Insurance Industry Sep 1, 2026 Authors Unstructured In this article Join our newsletter to receive updates about our features. In this article Structu…

Making the leap to specialized intelligence Join us for our inaugural conference, Forge 2026 Blog Making The Leap To Specialized Intelligence Making the leap to specialized intelligence PUBLISHED 9/10/2026 Table of Cont…

Introducing the Life Sciences Verification Program Sep 17, 2026 Today, we are introducing the Life Sciences Verification Program (LSVP), which gives life science professionals access to our Mythos, Opus, and Sonnet mode…

A team of three independent security researchers at Hacktron says it took less than 72 hours for them to hack into OpenAI employee accounts using Anthropic's Claude Opus 4.8 and 5, the Wall Street Journal reports. They were able to access OpenAI's GitHub repository, called "Monorepo," which reportedly contains "OpenAI's algorithmic secrets," according to the Wall Street Journal's sources. They stopped short of accessing internal code in Monorepo themselves, but sent a pull request from an employee's Codex account to prove they gained access. They were able to get in through Discourse, the third-party service that hosts OpenAI's community f … Read the full story at The Verge.

The trend is clear: open-weight models are taking an increasingly large bite out of production AI usage. On Monday, The The post Open-weight models now handle a majority of tokens on Vercel’s AI Gateway. But Anthropic still takes 64% of the spend. appeared first on The New Stack.

RSI or Recursive Self-improvement has been the talk of the town lately. The term came into surface when it was emphasized as the next step in the LLM evolution cycle by pioneers of the field like Sam Altman, Dario Amodei, and Elon musk. But also, via a research paper outlining the method titled: The Last AI Built by Humans. These two alone should help you realize […] The post Recursive Self-Improvement: The Last AI Built by Humans appeared first on Analytics Vidhya.

arXiv:2609.19315v1 Announce Type: new Abstract: Large language models (LLMs) provide a flexible interface for long-horizon robot planning, but generated plans often fail to respect embodiment constraints, recover from planning errors, or reason effectively under partial observability. We present GAVEL, a framework for verifying and repairing long-horizon LLM planning built around an explicit graph world model. The graph represents relevant object-relations, action pre-conditions and effects, and probabilistic beliefs over unobserved object locations. This model can predict the consequences of LLM-generated actions before execution, detect violations, and repair those whose corrections follow directly from the world model. This method also reserves LLM replanning solely for errors requirin…

arXiv:2609.19302v1 Announce Type: new Abstract: Open datasets describing human behavior in environments shared with mobile robots remain limited, particularly for retail activities that combine locomotion, reaching, object handling, and robot guided movement. This paper introduces OHRID Retail, an open, human centered multimodal dataset collected from 16 healthy adults performing a simulated shelf picking task under three within participant conditions: no robot, low speed robot guidance, and high speed robot guidance. Each participant completed two trials per condition. Whole body kinematics were recorded using 17 Xsens Awinda inertial sensors and muscle activity was measured at 10 locations using Delsys Trigno surface electromyography sensors. Descriptive analyses demonstrate variation i…

arXiv:2609.19216v1 Announce Type: new Abstract: Research on 4D millimeter-wave radar perception algorithms has flourished in recent years, extending from signal processing and object detection to semantic segmentation, motion estimation, occupancy prediction, and dynamic scene reconstruction. This review organizes the field according to the evolution of perception tasks and algorithms. It first introduces radar fundamentals, data representations, and quality-enhancement methods, and then reviews object-level perception, motion and localization, local and dense spatial perception, and dynamic scene understanding. Across these directions, we compare radar-only learning, multimodal fusion, and cross-modal supervision and knowledge distillation. Particular attention is paid to how elevation,…

arXiv:2609.19200v1 Announce Type: new Abstract: Underwater visuomotor policy learning has focused primarily on single manipulators, while bimanual imitation learning has been studied largely in air. We present ULOHA, an underwater bimanual robot learning platform that combines custom-designed leader--follower hardware with software extensions to LeRobot, integrating teleoperation, multi-view sensing, demonstration collection, policy training, and autonomous deployment. Real-robot experiments demonstrate a range of coordinated underwater bimanual behaviors, including inter-arm transfer, shared-object manipulation, and buoyancy-driven interception. We evaluate ACT, Diffusion Policy, and the vision--language--action model SmolVLA on the platform. We investigate how learning methods and execu…

arXiv:2609.19451v1 Announce Type: new Abstract: The Mobile Unified Multimodal Understanding (MUMU) Challenge requires a single efficient model to jointly perform multi-concept image tagging, open-vocabulary object detection, and image captioning. We present Efficient Unified Multimodal Understanding (EUMU), the winning solution for the MUMU Track of the 8th LSVOS Challenge. EUMU builds on a shared pretrained multimodal model, using its prompt-based capabilities for detection and captioning and training lightweight heads on shared visual features to predict quality, scene, and event tags. Rather than treating the three tasks independently, EUMU applies task-aware inference refinement by reusing task outputs as cross-task cues. For detection, caption cues help recover objects missed by the…

arXiv:2609.19444v1 Announce Type: new Abstract: Fine-grained evaluation of glomerular pathology must distinguish normal glomeruli from abnormalities such as global and segmental glomerulosclerosis, obsolescent, ischemic, solidified, disappearing, and atubular glomeruli. Supervised classification requires labeled examples of every category, which is impractical when subtypes are rare or absent from the training cohort. One-class anomaly detection offers an alternative by modeling normal data and scoring deviations, allowing previously unseen abnormalities to be detected. We use the frozen residual U-Net backbone of Omni-Seg, pretrained to segment structurally normal renal primitives without abnormal-subtype labels. We propose NoRDeC (Normal-Reference Detection and Characterization), a fram…

arXiv:2609.19421v1 Announce Type: new Abstract: Gaussian Splatting has significantly improved the quality of novel view synthesis with explicit Gaussian representation. However, we observed that existing 3D Gaussian Splatting methods (3DGS) often suffer from surface collapse issues on reflective regions, and thus produce inferior geometry and low-quality specular. In this work, we propose a physically-based deferred rendering framework, named Reflection-aware Gaussian Splatting (RGS), that can accurately model specular regions and improve novel view synthesis performance. Specifically, we found that a powerful 3D foundation model can provide a strong 3D geometric prior to foster correct geometric modeling. Based on this, we propose a cross-view shape consistency regularization to regulari…

arXiv:2609.19393v1 Announce Type: new Abstract: Work zones alter lane geometry through temporary traffic controls and closures that may be absent from on-board maps, challenging autonomous vehicle (AV) perception and planning. Generalization is also limited by scarce public datasets with structured geometric supervision. We present WorkZonePlan, a dataset comprising 149K+ synthetic and 5K+ real-world multimodal samples with 3D annotations for lane boundaries, work zone boundaries, and driving trajectory options. It also provides 76 closed-loop CARLA scenarios replayed under three weather conditions, yielding 228 Bench2Drive-format evaluation routes. We introduce WAVE (Work-zone-focused AV data generation in Virtual and rEal Environments), a semi-automated pipeline for creating the dataset…

arXiv:2609.19384v1 Announce Type: new Abstract: Scaling deep learning faces critical bottlenecks: data exhaustion, exponential training costs, and resource concentration. Model merging combines pre-trained checkpoints without gradient descent, offering orders-of-magnitude savings versus retraining. Combining independently trained vision models is difficult when their architectures and parameter shapes differ. Existing weight-space merging methods generally assume aligned, shape-compatible checkpoints, whereas a Vision Transformer (ViT) and a state-space model (SSM) implement token mixing with different operators. We study a hybrid Heterogeneous merging setting that retains both architectures while aligning parameter groups by semantic role. Our proposed Riemannian--Lorentz Parameter Fusio…

arXiv:2609.19358v1 Announce Type: new Abstract: Three-dimensional object detection for autonomous driving is dominated by detectors trained on large corpora of human-annotated 3D boxes. Such a detector learns a fixed category list, and everything outside it is invisible. This paper asks whether the task can be solved training-free and open-vocabulary. A promptable segmentation model (SAM3), queried with class names as text prompts, supplies instance masks in the vehicle's six surround-view cameras, and the masks are turned into metric 3D boxes using the geometry of the scene. The core is a controlled three-stage comparison on nuScenes in which 2D detection is held fixed and only the source of 3D geometry changes. Geometry predicted from images alone reaches 0.183 mean average precision (m…

arXiv:2609.19354v1 Announce Type: new Abstract: Automated action quality assessment (AQA) in Olympic sports remains a challenging task due to the complexity of human motion and the subjectivity inherent in expert judging. This work evaluates the capability of open-source Vision-Language Models (VLMs) to perform zero-shot action quality assessment on Olympic diving videos using the AQA-7 benchmark dataset. In this regard, a regression-based framework is pro-posed to leverage both the semantic reasoning and phase-level sub-scores generated by the VLMs, combining TF-IDF vectorization, dimensionality reduction, and ensemble learning to predict final competition scores. Experimental results show that standalone VLMs achieve moderate Spearman correlations below 0.32, while the proposed ensemble…

arXiv:2609.19230v1 Announce Type: new Abstract: Ultrasound is the most widely deployed imaging modality worldwide, yet clinical AI remains fragmented into narrow single-task models that fail when device, operator, or anatomy changes. Here we present SonoCorpus, an open resource unifying 456,963 images and 1,626,085 expert masks from 53 public datasets spanning 24 clinical applications and 17 countries, and SonoBase, an interactive segmentation foundation model pretrained on it. Across fifteen evaluation datasets introducing new organs, devices, operators, and geographies, SonoBase outperforms SAM2, MedSAM2, and the concept-promptable MedSAM3 on every dataset and matches per-dataset specialist models trained on the same data; on fully external data it exceeds the accuracy these baselines a…

arXiv:2609.19158v1 Announce Type: new Abstract: Knowledge graphs are usually integrated into question answering by encoding a retrieved subgraph with a graph neural network and fusing it with the language model in the online inference path. The same subgraph is therefore re-encoded from scratch every time a pair is scored, across training epochs, seeds, and evaluation runs, even though the knowledge graph never changes. We ask whether the retrieved knowledge graphs can instead be compiled once, offline, and then accessed as read-only memory. VisKG-LM shows that it can, by decoupling graph encoding from language reasoning. It serializes each retrieved candidate-specific subgraph as Relation-Labeled Paths and renders the result as an image whose two-dimensional layout preserves the branchin…

arXiv:2609.19156v1 Announce Type: new Abstract: Data-driven fine-tuning is widely adopted to enhance reasoning in Large Language Models (LLMs) due to its simplicity and efficiency. However, mainstream imitation learning methods that rely exclusively on perfect reasoning trajectories suffer from a Scaling Collapse: when the problem set is limited, increasing positive examples fails to yield continuous improvement. However, during inference, an LLM can not guarantee that every intermediate step is correct and is therefore prone to errors. Once such errors arise, the LLM often struggles to recover and may be further misled by the accumulation of previous mistakes. To address this, we propose Reflective Recovery, a simple yet effective self-supervised approach that transforms failed reasoning…

arXiv:2609.19155v1 Announce Type: new Abstract: In Human-LLM dialogue, follow-up user utterances may implicitly conflict with earlier intents, leading the LLM to misinterpret user needs and generate inappropriate responses. A reliable dialogue system should proactively detect user-side conflicts before generating a response and seek clarification when necessary. However, prior work has largely focused on LLM-side conflicts, leaving user-side conflicts underexplored. To fill this gap, we construct UC-Bench, a human-annotated benchmark for evaluating user-side conflict detection. Preliminary experiments show that existing LLMs struggle with this task, especially when conflicts arise from implicit incompatibilities grounded in dialogue history. To improve lightweight LLMs with limited traini…

arXiv:2609.19152v1 Announce Type: new Abstract: Misinformation detection tools often rely on binary true and false classifications or models trained on historical examples, limiting their usefulness when novel misleading narratives emerge. Here, we present FakeSpotter, a content- and strategy-agnostic tool designed to estimate the viral misinformation risk of textual content by measuring structural fingerprints of misinformation rather than directly adjudicating truthfulness. FakeSpotter operationalizes a theory-driven framework across linguistic, narrative, logical, and critical-thinking dimensions, using repeated LLM assessments and domain-specific logistic regression classifiers for short and long texts. In a labelled corpus of 764 texts from social media and FakeNewsNet, FakeSpotter a…

arXiv:2609.19150v1 Announce Type: new Abstract: Large language models (LLMs) encode rich stylistic structure in their hidden activations, but discovering which stylistic dimensions are salient for a given prompt typically requires supervised contrastive data. We present a training-free, prompt-conditional alternative: we repeatedly sample completions of a single prompt at elevated temperature, apply Principal Component Analysis (PCA) to the pooled hidden activations, and label the resulting axes automatically from the pole generations. We validate the discovered axes against 245 human-elicited stylistic annotations in a two-phase study. On our strongest model (Qwen-3.5-4B-Instruct), the top two axes match spontaneously requested human dimensions with 72.8% precision and 43.6% macro-recall…

arXiv:2609.19149v1 Announce Type: new Abstract: Subliminal learning shows that language models can transmit a hidden trait through outputs that appear unrelated to it. One proposed explanation, token entanglement, links animal and number tokens through the model's output vocabulary. Yet existing measurements answer different questions: whether outputs co-vary, fixed output vectors align, an answer can be read from a hidden state, or that state causally controls the answer. We measure each separately in a fixed animal-number prompting protocol. From Llama-3.1-8B to 70B, fixed output-vector similarity predicts behavior less well: the paired mean correlation change is -0.080 (95% CI [-0.127, -0.035]). A fixed output-head readout shows no resolved change in normalized depth AUC. To test contr…

arXiv:2609.19148v1 Announce Type: new Abstract: Ambivalence and hesitancy (A/H) are affective states in which individuals express contradictory signals across facial, vocal, and linguistic channels. Automatically recognising A/H in clinical videos requires detecting cross-modal disagreement -- the signal that standard fusion methods suppress. Based on the conflict-aware multimodal fusion framework of Bekhouche et al., we present the Modality Discrepancy Transformer (MDT). MDT enriches the original 6-token design to a 9-token representation comprising three modality embeddings, three absolute-difference features, and three Hadamard-product discrepancy features learned through linear projections. These nine tokens undergo Transformer self-attention, with FiLM-based text-conditioned modulati…

arXiv:2609.19363v1 Announce Type: new Abstract: The query and key projections $\WQ,\WK$ in attention are almost always trained by Euclidean optimizers with no constraint on their geometry. We constrain them to the Stiefel manifold and optimize them there with a Riemannian Adam that carries one scalar second moment per frame, caps its step by a trust region, and retracts polarly. Four propositions prove this update is steepest descent in the embedded metric, independent of gradient scale, well conditioned, and exactly $\mathrm{O}(d)$-equivariant, each certified numerically in \texttt{float64}. A fifth supplies the mechanism: weight decay has \emph{identically zero} Riemannian gradient on $\St(d,r)$, since $W = W I_r$ lies in the normal space, so the learned attention geometry survives the…

arXiv:2609.19356v1 Announce Type: new Abstract: We introduce a new method to guide flow matching models. Our approach, which we call probe guidance, uses the frozen internal states of an existing diffusion model to construct a guidance signal. This works using a similar principle as autoguidance, but eliminates the need for an additional forward pass at inference time and provides a reliable path to ensure that the weak and strong model share similar dynamics. We apply and benchmark this method on continuous diffusion language models, where probe guidance sets a new state-of-the-art performance on unconditional generation. When applied to a 1.7B diffusion language model, probe guidance consistently improves on multiple choice question answering benchmarks. Using our probes, we study the t…

arXiv:2609.19242v1 Announce Type: new Abstract: Block diffusion language models (BDLMs) combine autoregressive dependencies across blocks with parallel denoising within blocks, but long-context training is constrained by distributed attention communication and activation memory. Conventional context parallelism (CP) shards the combined clean-plus-corrupted sequence by position, communicating shared clean K/V together with block-specific corrupted K/V and their gradients. We observe that the BDLM objective separates over target blocks. We introduce block parallelism (BP), a new distributed parallelism dimension that assigns each corrupted-block computation to one rank. To scale BP to long contexts, we introduce context-sharded block parallelism (CSBP), which also shards the shared clean se…

arXiv:2609.19213v1 Announce Type: new Abstract: In this paper, we introduce layer-wise curriculum learning for efficient LLM compression. The proposed method facilitates the knowledge transfer from the teacher model to the student model, utilizing a curriculum learning approach that begins with easier optimization tasks and progressively tackles harder ones. In order to adopt the layer-wise learning in LLM compression, we partition the whole model into multiple segments consisting of layers, thereby enabling more computationally efficient knowledge transfer for LLMs. Based on our theoretical analysis of cumulative error phenomenon, layer-wise curriculum learning accelerates convergence while stabilizing the knowledge transfer process. In addition, we present a feature caching method with…

arXiv:2609.19209v1 Announce Type: new Abstract: Generative query suggestion aims to enhance user engagement by anticipating user intents and recommending relevant follow-up queries. A central challenge is to generate slates whose individual queries are useful while the slate covers distinct intents. We propose an Intent-Driven Query Suggestion Framework with dual-stage optimization. First, intent-aware diversity modeling constructs intent-aligned supervised fine-tuning (SFT) data and uses an Intent-Aware Diversity Reward to optimize intent coverage. Second, query-level credit assignment routes individual quality signals to the corresponding query tokens while sharing a slate-level diversity signal across the slate. Experiments on a large-scale production dataset, including online A/B test…

arXiv:2609.19212v1 Announce Type: new Abstract: Systematic generalization, the ability to solve novel problems by recombining known atomic elements, is central to human intelligence but difficult to study rigorously under controlled settings. Existing studies therefore rely on simplifications such as approximately linear action composition, productivity-based tests, and action-explicit goals, which make systematic generalization easier to study but omit some essential aspects of this capability. To characterize what these simplifications miss, we adopt a reasoning-centered lens and introduce TranSGrid, a testbed that brings deductive, inductive, and abductive reasoning together within a unified task. Experiments with seven Transformers on 4,800 TranSGrid instances show that all models per…

arXiv:2609.19203v1 Announce Type: new Abstract: AI applications have shifted from single, monolithic foundation models (FM) to compound agentic systems. Yet today's stacks remain fragmented: even as protocols (e.g., MCP, A2A) ease tool/agent connectivity, each framework embeds an implicit runtime for state, memory, budgets, and guardrails, making behavior non-portable and governance brittle. It mirrors computing before operating systems, when every program re-implemented basic services. This position paper argues that the field now needs a Foundation Model Operating System (FMOS) -- a system layer that virtualizes FM interactions analogous to how virtual machines abstract physical hardware, giving applications the illusion of dedicated, trustworthy FM instances with effectively unbounded…

arXiv:2609.19182v1 Announce Type: new Abstract: Benchmarks are central to how progress in large language models (LLMs) is assessed and communicated. Yet model rankings alone reveal little about how evaluation requirements themselves are changing. The expanding variety of benchmarks offers another perspective: what researchers expect LLMs to do, and what they count as successful performance. We systematically map 14,767 papers introducing or updating evaluation resources from arXiv submissions between January 2022 and August 2026. Using staged screening and automated full-text coding, we examine changes in target systems and domains, evaluation materials and conditions, and scoring mechanisms. The collection shows growing emphasis on action, interaction, and professional applications, whil…

Open Source Agent Architecture LangChain Building a Harness with Jev September 17, 2026 5 min Go back to blog Create agents Agents run in a loop: an LLM decides what to do, a tool executes, a model evaluates the results…

Activation steering has emerged as a powerful method for guiding the behavior of generative models towards desired outcomes such as toxicity mitigation. However, most existing methods apply interventions uniformly across all inputs, degrading model performance when steering is unnecessary. We introduce Dynamically Scaled Activation Steering (DSAS), a method-agnostic steering framework that decouples when to steer from how to steer. DSAS adaptively modulates the strength of existing steering transformations across layers and inputs, intervening strongly only when undesired behavior is detected…

How To Write With An LLM Thomas Ptacek on using LLMs as copyeditors, not as writing assistants: Rule Number One: You may not use a single word an LLM suggests to you. [...] I think that as a form of intellectual personal protective equipment you should adopt the rule that any specific turn of phrase an LLM suggests is off limits. Be strict about the rule! I won't let LLMs write content for my blog, but I use them for fact-checking, spelling and grammar and as an occasional thesaurus (see my proofreading prompt). The rule to never use a turn of phrase suggested by an LLM feels good to me. The text has that weird smell to it, and it's also a good principle to help stay disciplined. Later in this piece Thomas shows a screenshot of his personal LLM copyediting tool (see also this Twitter thre…

The models support real-time conversation by enabling users to interact without the traditional prompt-response structure.

Regulating frontier AI could sideline smaller open source developers and push enterprises to take on more of the safety and governance burden themselves.

Discussion | Link

Discussion | Link

Discussion | Link

Akademy talk imagines Plasma assembling itself around a personal model of each user

Discussion | Link

In this article, you will learn how a vector database works under the hood by building one from scratch in ten incremental steps using Python...

Tech bosses have called for a slowdown in artificial intelligence and formal guardrails to be introduced by the US government. But President Trump is not moved by the doomsday predictions, calling them a ‘hoax'. What is behind his affection for the AI industry? Jonathan Freedland and Guardian US tech editor Blake Montgomery discuss Continue reading...

Liberal MP speaks of AI’s shortcomings as cyber chief tells inquiry the tech is needed to fend off ‘highly capable malicious cyber actors’ Get our new political email, free app or daily news podcast When a man with a terminal illness wrote to his local MP Andrew Hastie, telling him he planned on ending his own life with voluntary assisted dying, Microsoft Copilot suggested Hastie reply with “congratulations!”, “great to hear from you” or “that is wonderful news!”. Hastie shared the episode at a parliamentary inquiry on Friday to underline the shortcomings of the American-run software as he called for AI run by Australians. Continue reading...

Warwick Levy, who runs Lonely Kids Club, says AI may have been used to scrape his entire website and replicate designs Get our breaking news email, free app or daily news podcast After 15 years running a small business making T-shirts, Warwick Levy says it was “brutal” when he first discovered an identical design for sale on Temu. It turned out there was an overwhelming amount of T-shirts identical to his own for sale on the Chinese e-commerce marketplace. Between 2025 and earlier this year, Levy, 37, found thousands of rip-offs. Continue reading...

After a week of alarming warnings about the technology’s potential to destroy the world as we know it, our expert tech reporters will take your questions on the reality of the AI threat. Join us at 3pm BST (4pm CEST, 10am EDT) where they will answer live Sign up or sign in to ask your question The last week or so has seen disturbing claims about the capacity of superintelligent AI to destroy the planet. It has also seen warnings coming from figures within the industry itself, including Anthropic’s report that criminals, state-sponsored groups, spyware vendors, scientists and propagandists have attempted to use its models to “design missiles and bombs, create deadly pathogens and surveil dissidents” and Elon Musk’s claim it could be “more dangerous than nukes”. Our reporting team has been…

Discussion | Link

Tech bosses have called for a slowdown in artificial intelligence and the creation of formal guardrails from the US government. But why is President Trump unmoved by the doomsday predictions, dismissing them as a ‘hoax’? What lies behind his embrace of the AI industry? Jonathan Freedland speaks to the Guardian US tech editor, Blake Montgomery Watch Jonathan Freedland on our new Politics Weekly YouTube channel Send your questions and feedback to [email protected] Listen to the latest season of Black Box Continue reading...

Discussion | Link

Discussion | Link

Discussion | Link

Discussion | Link

Discussion | Link

Discussion | Link

Discussion | Link

Discussion | Link

Discussion | Link

A significant portion of enterprises struggle to demonstrate clear ROI from AI initiatives, but some have found areas where AI can provide productivity gains and even revenue growth.

Discussion | Link

Discussion | Link

Discussion | Link

Discussion | Link

Andy Burnham’s focus on immediate domestic problems leads some to fear issue has dropped off government’s radar Towards the end of Keir Starmer’s time in office, his senior ministers, alarmed by the latest developments in artificial intelligence, began drawing up plans for a new AI safety law. They ordered a review of existing legislation to see what powers they already had, according to those briefed on the plans, and were exploring whether they could force the world’s most advanced technology companies to submit their products for safety testing before launching them. Continue reading...

As AI gets more capable, enterprises are running into a different set of problems: managing models, data, permissions and governance.

OpenAI introduces the Australian Youth Safety Blueprint, a six-pillar roadmap for safer AI experiences that protect and empower young people.

Most people think AI is risky, but they don’t agree what to do next.

We are expanding our AI & Economy team with world-class academic advisors, fellows, and core internal researchers.

On the morning of June 9th, Laura Lin was working from her home in Lanesville, a rural southern Indiana town about 15 miles from the Kentucky border. She was on a Zoom call, unaware that the heavy rain outside was beginning to flood her yard. "I look over to where the barn is over there, and I see pieces of my wood floating, and I was like, 'What?' And I immediately was like, 'I have to go.' Close my laptop, and I get my kids up, and I'm like, 'Something's wrong,'" Lin recalls. Lin and her family got out safely and sheltered at a neighbor's house, but Lanesville got over 8 inches of rain within just a few hours that day, way over the thre … Read the full story at The Verge.

We are used to the idea that our fellow humans might intentionally mislead or manipulate us, but the idea that machines can now do the same is deeply unsettling. Researchers are racing to find solutions before it’s too late. By Snigdha Poonam. Read by Maya Saroya Read the text version here Support the Guardian today: theguardian.com/longreadpod This article was supported by a grant from the Tarbell Center for AI Journalism Continue reading...

arXiv:2609.19272v1 Announce Type: new Abstract: Research in humanoid robotics has achieved rapid progress in locomotion, and recent results have pushed the boundary on autonomous navigation. We demonstrate that a standard single-stage RL navigation pipeline struggles to scale to multi-level and multi-story terrain, limited by the difficulty of complex humanoid terrain interactions such as stairs. To overcome this challenge, we decompose the navigation problem into two pieces. First, we train a policy operating on the reduced order dynamics but with full 3D LiDAR observations to navigate complex, multi-story terrain. We then utilize this navigation knowledge to kickstart a policy operating on the full-order humanoid dynamics, with a frozen locomotion policy in the loop. Additionally, we de…

arXiv:2609.19228v1 Announce Type: new Abstract: Dexterous robot hands must often grasp objects whose shape, size, and pose are known only approximately. Grasp planners typically require accurate object models or correct errors with feedback, but how much inaccuracy a closing motion can tolerate on its own remains unclear. To address this question, we designed a motion planner based on four principles: a coarse template of the object, human grasp types, an object-centric interaction, and compliant, sliding contacts instead of prescribed contact points. This paper presents the planner, implemented through virtual model control, and its evaluation on a Shadow Dexterous Hand. Without feedback, the planned closing motions tolerated size errors of about 1cm and pose errors of several centimeter…

arXiv:2609.19204v1 Announce Type: new Abstract: Robotic systems operating in dynamic environments require visual perception that evolves continuously with the incoming sensory stream. Event cameras provide microsecond temporal resolution and asynchronous sensing, but most learning-based methods accumulate events into frames or temporal bins, introducing an integration delay that can limit fast reaction. Here we propose REACT, a fully spiking state-space model for event-driven temporal perception that processes raw events one by one, without temporal accumulation. REACT uses a complex-valued spiking neuron, C-SiLIF, whose continuous-time dynamics are driven by the physical inter-event interval, allowing its internal state to evolve at the temporal resolution of individual events. We evalua…

arXiv:2609.19196v1 Announce Type: new Abstract: Collecting data for manipulation with high-DOF hands is challenging, as interfaces must capture rich hand motion while rendering the contact interactions essential for precise manipulation. Existing data collection approaches face a trade-off: teleoperation ensures deployment consistency but lacks force feedback, while handheld (in-the-wild) systems provide natural force transparency but introduce a visual embodiment gap at deployment. We present DITTO, a Dexterous Interface for Transparent TeleOperation, which resolves this through the anatomically informed co-design of a dexterous 7-DOF robotic hand and a kinematically equivalent motorized exoskeleton. A 1-to-1 actuator mapping between the exoskeleton and robotic hand enables handheld (in-…

arXiv:2609.19236v1 Announce Type: new Abstract: Objective: Incomplete navigation of anatomy during ureteroscopic kidney stone surgeries can contribute to repeat interventions. While skilled surgeons have lower reintervention rates, there are no objective metrics to quantify scope-navigation performance to evaluate when a trainee becomes skilled. This work aims to recover ureteroscope trajectories from endoscopic video and derive navigation metrics to quantify differences in skill. Methods: We propose RAUL, a reference-assisted reconstruction framework for recovering ureteroscope trajectories from ureteroscope videos only in phantoms. For each phantom, we use a slow, high-quality reference exploration video to generate a reference reconstruction. We localize subsequent exploration videos a…

arXiv:2609.19154v1 Announce Type: new Abstract: While Large Language Models (LLMs) achieve high accuracy on established Classical Chinese Poetry benchmarks, it remains challenging to distinguish transferable Linguistic-Aesthetic Reasoning from reliance on familiar pre-training patterns. To address this issue, we introduce Neo-Classic, an evaluation benchmark that combines a constructionist Out-of-Sample (OOS) dataset with a suite of reverse understanding probes. Unlike traditional benchmarks that rely on verification or generation over historical corpora, Neo-Classic comprises strictly metrical poetry authored by contemporary experts, reducing the possibility of direct retrieval. We evaluate state-of-the-art models, including Qwen3-Max, Gemini-3-Pro, and DeepSeek-V3.2, across five behavio…

arXiv:2609.19153v1 Announce Type: new Abstract: Empirical legal scholarship increasingly treats judicial text as data, and much of it still runs on sparse, interpretable pipelines -- TF-IDF features and linear classifiers -- because the textual feature is often the object of study, not merely a means to a prediction. Yet these pipelines inherit a chain of preprocessing defaults from mid-century information retrieval that were never validated against classification accuracy, the most entrenched being stopword removal. This study introduces an exhaustive single-word ablation that measures a preprocessing step's effect directly against the downstream objective, and applies it to stopword removal as the hardest case to dislodge. Matching Supreme Court Database labels to Caselaw Access Project…

arXiv:2609.19151v1 Announce Type: new Abstract: Generative AI (GenAI) applications have achieved rapid consumer adoption, yet little large-scale research examines user-perceived quality, trust, and adoption barriers. We present one of the first cross-application analyses of app store reviews for six major GenAI applications (ChatGPT, Gemini, Microsoft Copilot, Claude, DeepSeek, and Perplexity), comprising 17,012 English-language reviews from Google Play and the Apple App Store. We combine BERTopic topic modeling with RoBERTa sentiment classification and evaluate cross-application differences using chi-square, Kruskal-Wallis, and multinomial logistic regression with Bonferroni correction. Both components are validated against human coding using a stratified sample of 300 reviews. Results s…

arXiv:2609.19337v1 Announce Type: new Abstract: We present Personalized Federated Hierarchical Gaussian Processes (pFedHGP) for probabilistic regression and classification when data are distributed across heterogeneous clients. Each client's latent function decomposes into (i) a shared global component, (ii) a client-specific deviation that shares the global kernel structure, and (iii) a flexible local residual. Sparse inducing-variable approximations and federated variational inference keep raw data local while the server synchronizes only low-dimensional statistics for the shared component. Full predictive distributions support uncertainty-aware decisions. In application studies, pFedHGP attains perfect fault classification in press tonnage monitoring using 13.77% of labeled cycles and…

arXiv:2609.19288v1 Announce Type: new Abstract: Learning in recurrent neural networks can fundamentally reshape their underlying dynamics, transforming initially chaotic activity into stable task-dependent behavior. We develop a non-equilibrium dynamical mean-field theory(DMFT) to describe this transition during learning. We show that a slow feedback-driven learning process generates an evolving effective feedback strength that drives the network through a transition from chaotic to stable dynamics defined by a bifurcation of the DMFT solution. By deriving the two-time correlation function throughout learning, we identify a critical feedback strength and a corresponding learning rate dependent critical time separating these regimes. The transition arises from the progressive deformation o…

arXiv:2609.19279v1 Announce Type: new Abstract: Running artificial intelligence (AI) models directly on edge devices such as smartphones, wearables, and drones offers low latency, pervasive scalability, and data privacy, but these devices rarely carry the computing capability that modern neural networks demand. Edge accelerators have been developed in response, yet each adds computing hardware to devices already constrained in size, weight, power, and cost (SWaP-C). An alternative lies in what these devices already carry: the frequency mixer in every wireless radio multiplies signals in time, natively performing convolution in the frequency domain. Here we introduce radio-frequency convolutional neural networks (RF-CNNs), which repurpose existing communication hardware for CNN inference.…

arXiv:2609.19243v1 Announce Type: new Abstract: Spectral co-clustering is a useful tool for discovering latent structure in word-document matrices, but its reliance on singular value decomposition (SVD) can make standard formulations expensive on high-dimensional data. This paper presents two randomized approximations for normalized spectral co-clustering of bipartite text data when the numbers of document and word clusters may differ. The first method uses randomized SVD through random projection, while the second combines partial SVD with element-wise random sampling. Across real-world and synthetic datasets, both methods reduce runtime relative to the full-SVD baseline, but their behavior depends on matrix sparsity. The random projection method is the more reliable approximation across…

arXiv:2609.19448v1 Announce Type: new Abstract: In both cognitive science and computer science, goals are conceptualized as cognitive states that flexibly combine with world knowledge to organize and specify purposeful behavior. In this way, goals are compositional representations whose content relates to rational behavior. We here draw attention to goals as representations and their content because it highlights a parallel with other areas in cognitive science - in particular, the syntax-semantics interface in linguistics and logic - while also foregrounding foundational questions about the expressivity, design, and efficiency of different goal representations. For example, goals are typically taken as fixed and imposing constraints on desirable behaviors, but we can also identify constr…

arXiv:2609.19180v1 Announce Type: new Abstract: Language models face unique challenges in analyzing interdisciplinary scientific research literature. In biophysics research, faithful answers require grounding observed data in source evidence, interpreting it through a quantitative physics model, and linking it to a biological mechanism. To address this challenge, we introduce BioPhys-Bridge, a novel benchmark dataset for evidence-grounded scientific reasoning over biophysical literature. Each case contains evidence blocks, stable evidence IDs, quantitative values, units, equations, assumptions, mechanisms, and next decisions as grounding targets for question answering (QA) and retrieval-augmented generation (RAG). The initial release contains 500 cases, 1,517 agent-facing tasks, and cover…

arXiv:2609.19170v1 Announce Type: new Abstract: Emphatic temporal-difference learning (ETD) stabilizes the expected off-policy TD update and changes its projection geometry, but neither property determines constant-stepsize sampled dynamics. We construct an ergodic two-state counterexample in which the ETD mean map contracts while the sampled product has a positive top Lyapunov exponent. Regenerative-cycle analysis separates this sign from the infinite variance of the follow-on trace. We introduce regularized emphatic TD (RETD), a normalized first-order post-shock repair that leaves the trace and importance ratios unchanged, stores the emphatic TD signal in a leaky scalar state, and releases a delayed correction. RETD's raw equilibrium is an affine shift of the ETD equilibrium; single- an…

← Blog VQ-bench: a Composable Vector Quantization Framework Ashwin Padaki, Amir Ingber, Edo Liberty Sep 17, 2026 EngineeringResearch Share: Before a vector database can search vectors, it has to store them. But storing…

Amazon SageMaker HyperPod Inference Gateway is a Kubernetes-native, GPU-aware routing add-on for Amazon EKS. It uses real-time GPU signals to send each inference request to the best-suited pod, cutting first-token latency by up to 82% with no changes to your model servers or client applications.

Microsoft's AKS engineering team open-sourced TauGrid on August 28, 2026, packaging the tau CLI, Kueue queueing, KubeRay orchestration, GPU node health monitoring and observability into one Helm install. It is MIT licensed and deployable now on any Kubernetes 1.30+ cluster with GPU nodes, kubectl and Helm 3.0 or later. The post Microsoft Open-Sources TauGrid: A Kubernetes-Native Stack for GPU AI Workloads appeared first on MarkTechPost.

US cybersecurity researchers who conducted hack say ‘scope of what we could theoretically access was huge’ Cybersecurity researchers have hacked into OpenAI with the help of Anthropic’s Claude chatbot, in the latest example of security issues at the company. A team at a US-based startup compromised multiple OpenAI employees’ ChatGPT accounts, starting a process that enabled them to access their target’s software cache – and potentially more. Continue reading...

Amazon Connect Talent is an AI hiring solution built for talent acquisition leaders managing scaled hiring. It delivers AI-led interviews, data-driven assessments, and consistent evaluation, helping recruiters identify strong candidates more efficiently while providing applicants with a flexible interview experience. Informed by decades of Amazon's hiring science, Amazon Connect Talent provides transparency for every assessment, interview, and candidate score, enabling recruiters to stay in control of final hiring decisions.

arXiv:2609.19328v1 Announce Type: new Abstract: In recent years, aerial manipulation has attracted increasing attention as a key to expand the application of aerial robots. In this work, we focus on two major research directions for achieving versatile aerial manipulation: (i) acquiring high environmental adaptability using soft manipulators, and (ii) expanding the feasible wrench space by distributing thrusters along the manipulator. However, no aerial robot has simultaneously satisfied these two requirements. Therefore, in this paper, we propose a morphing rotor-distributed aerial robot with flexible continuum links that achieves both high shape adaptability and an expanded wrench space. The flexible continuum links function as soft manipulators, passively conforming to the shape of the…

arXiv:2609.19194v1 Announce Type: new Abstract: AthenaZero is a bimanual manipulator designed to minimize inertia without compromising control authority. By utilizing quasi-direct drive actuation and transmission remotization techniques, the system achieves an effective endpoint mass comparable to that of a human---about an order of magnitude less than conventional robot manipulators. This characteristic, combined with its inherent torque transparency, makes AthenaZero exceptionally well-suited for dynamic manipulation. We describe the methodology} that led to this design and demonstrate the robot's capabilities on three baseball-inspired tasks: throwing, catching, and batting, which showcase complex interactions on human-comparable timescales where milliseconds matter. AthenaZero was cap…