424 points, 156 comments
337 points, 197 comments
Official technical announcement and publication from Oxford (AIDC) covering Deputy HR Manager.
Official technical announcement and publication from Oxford (AIDC) covering HR Officer.
General Science
Soft materials remember their deformation history, and identifying that memory from experiments is essential for predicting how these materials behave under real-world loading conditions. Chirp rheometry has recently emerged as a way to accelerate this characterization, compressing hours of conventional measurement into seconds and yielding thousands of stress-strain pairs per experiment. That den
Parallel reasoning improves the accuracy and robustness of large reasoning models by exploring multiple solution paths, but its computational cost grows with reasoning depth and branch count. Existing methods for managing these parallel paths typically rely on final-answer consensus, local token confidence, or isolated intermediate probes. However, these signals are often delayed, weakly tied to a
AI is reshaping cybersecurity for attackers and defenders alike. Learn how OpenAI is strengthening its defenses and what security teams can do now.
OpenAI joins PORTS-Pike project, expanding community investment and supporting thousands of Southern Ohio jobs
OpenAI funds 14 independent projects exploring new AI policy ideas to expand economic opportunity and strengthen societal resilience in the Intelligence Age.
250 points, 544 comments
We ran 904 DeepSWE rollouts on DeepSeek V4 Pro 0813 and Claude Fable 5. Fable leads pass@1 at 90x the cost; Pro wins pass@4, and a Pro-first cascade hits 82.7%.
Shadow traffic proves a candidate is operationally sound. It can't tell you if users like it better. Run the split at the endpoint instead of in your app code.
Official Groq (LPUs) technical update and publication covering Read more >.
Official LMSYS Chatbot Arena release and benchmark update covering Advanced CUDA Graph Techniques in SGLang.
484 points, 297 comments
LiDAR scene completion is a key component of 3D perception in autonomous driving, where the scene must be completed in real time to be usable in downstream tasks. Existing approaches typically follow an initialize-and-refine paradigm, in which a coarse initialization of the scene is first constructed, then refined into complete 3D geometry. Generative models are slower because they iteratively ref
Long-horizon robot manipulation requires a robot to both execute individual skills reliably and sequence them coherently over extended tasks. Most hierarchical vision-language-action (VLA) models make each such decision with a single forward pass, leaving no mechanism to allocate additional computation to difficult or consequential choices. We introduce ฯ_0-VLA, a hierarchical robot foundation mod
Methods for improving knowledge use in large language models typically fall into two regimes. Non-parametric retrieval offers flexible access to external knowledge, but adds retrieval latency, context overhead, and only shallow integration with the backbone. Parametric adaptation is efficient at inference time, but entangles knowledge with model weights and can be hard to update, audit, or transfe
Frontier open-weight models are increasingly available, but serving them still largely assumes datacenter infrastructure. We present FreeToken, an edge-native MoE serving system that treats a personal machine not as a small GPU, but as a unified, elastic inference platform. FreeToken co-designs the full serving stack, including model layout and loading, expert residency, CPU--GPU execution, agenti
Hybrid-thinking multimodal large language models (MLLMs) allow a single model to alternate between deliberative thinking and latency-efficient non-thinking inference. Although these modes differ in reasoning budget, their delivered responses should satisfy the same user-facing standard. Correctness alone may not characterize this response quality; we therefore evaluate task accuracy and response-p
Graph neural networks are commonly described through family-specific equations whose notation obscures shared computations and structural differences. We introduce a common layer equation that represents covered architectures through seven components: an update domain, channel set, propagation bank, per-channel message maps, channel-fusion operator, ego/residual map, and update map. The central fa
As text-to-image generative models advance, they raise critical safety concerns, particularly the generation of Not-Safe-For-Work (NSFW) content such as violence and nudity, further exacerbated by red-teaming adversarial attacks. Existing defenses predominantly operate under white-box assumptions, relying on text encoder optimization, weight editing, or inference-time intervention, and fundamental
Existing image editing frameworks predominantly follow the training paradigm of text-to-image diffusion models. However, extending this paradigm to image editing highlights two inherent discrepancies, specifically, the insufficient attention to edit concept granularity and the training inefficiency caused by sparse supervision signals. To address these issues, we establish a comprehensive hierarch
Unified image restoration (UIR) aims to recover high-quality (HQ) content from low-quality (LQ) images with different degradations using a single model. Most recent methods adapt large pretrained text-to-image (T2I) latent diffusion models for their strong capacity and generative priors. However, the variational autoencoder (VAE) in latent T2I models may discard restoration-sensitive details, whil
Looped language models have shown promising results on reasoning benchmarks, yet their potential for agentic tool use remains largely unexplored. We study this question in compositional tool-calling settings, where models must coordinate multiple API calls, maintain intermediate state, and preserve dependencies across tool interactions. We evaluate native and retrofitted looped language models on
AI systems are increasingly capable of contributing to mathematical research. In research practice, frontier-model reasoning is a limited resource, and expert mathematical review is even more sharply constrained. Allocating these scarce resources well is therefore central to making AI-assisted mathematical discovery efficient. In most current AI-for-math workflows, human effort is concentrated at
Embodied agents are increasingly used to close the gap left by end-to-end policy models. Yet the agentic path has not realized closed-loop learning in physical execution: existing harnesses remain largely open-loop, following fixed skills during rollout and reflecting only after an episode completes. Such post-hoc reflection cannot govern execution as it unfolds, because physical interaction requi
On-policy distillation (OPD) transfers teacher capabilities by supervising trajectories sampled from the student's own policy, yet its generalization behavior remains poorly understood, as most studies evaluate OPD on a single domain and on benchmarks close to the training data. We present a controlled study that varies one generalization factor at a time, from in-domain distribution shifts to cro
Intentional radio-frequency interference from low-cost GNSS jammers increasingly threatens the accuracy and reliability of satellite-based positioning. Mitigating this threat requires not only detection but also robust waveform classification and characterization, direction-of-arrival inference for localization, and impact estimation on receiver performance under realistic operating conditions; al
332 points, 129 comments
The rapid adoption of large language models has enabled the development of clinical multi-agent systems (MAS) capable of integrating multimodal patient data and supporting increasingly complex clinical decision-making. However, the deployment of these systems in real-world healthcare settings raises critical ethical concerns related to safety, fairness, accountability, transparency, and patient tr
The rapid adoption of large language models has enabled the development of clinical multi-agent systems (MAS) capable of integrating multimodal patient data and supporting increasingly complex clinical decision-making. However, the deployment of these systems in real-world healthcare settings raises critical ethical concerns related to safety, fairness, accountability, transparency, and patient tr
Vision-language-action (VLA) models have become a dominant paradigm for generalist embodied agents, demonstrating strong complex and long-horizon task completion in structured settings. Yet it remains an open question whether current VLA systems can benefit from more effective architectural design, scale to substantially larger and more heterogeneous data regimes, and achieve broader generalizatio
Language-specific competency (LSC) is the phenomenon of a language model performing better or worse depending on the language of the prompt. In other words, a language model outputs different (and potentially incorrect) responses to the same semantic query when prompted in different languages. Prior work attributes this to an internal misalignment of semantic representation across languages. Curre
LLM-based agents can act on behalf of a user to access cloud services, call tools, or invoke agents. At session start, the agent's permissions are set but remain static, and each request is evaluated independently, without considering prior actions. Within its permissions, an agent may act contrary to the delegated task, combine individually permitted actions into a prohibited outcome, or delegate
We introduce TinyCast, an attention-free zero-shot forecaster that emits a predictive distribution from 146,505 parameters, on the premise that at this size the periodic structure of a context is worth computing rather than learning. A zero-parameter spectral detector supplies the dominant periods, the context is folded on their phase, and a dilated convolutional encoder and a block-autoregressive
632 points, 521 comments
Earth observation foundation models (EOFMs) are emerging as reusable representation frameworks for data-driven retrieval, prediction and process modelling within ecohydrology, which integrate EO, meteorological forcing and process models to characterise coupled water, energy and carbon dynamics in vegetation and soil across scales. However, there is yet to be an ecohydrology-specific synthesis ass
338 points, 203 comments
Memory is becoming core infrastructure for long-horizon LLM agents, yet existing evaluations offer limited guidance on which memory substrate, namely the underlying medium in which memory is represented and stored, should be used under different operating regimes. We present a controlled harness evaluation of memory substrates for memory-augmented agents, covering dense and sparse indices, text re
When a long-horizon agent execution fails, outcome-level evaluation reveals the unsuccessful result but not where the decisive error entered the trajectory. Developers must then inspect the full execution to identify the responsible role and localize the earliest decisive root-cause step. Existing failure-attribution benchmarks largely focus on shorter traces, leaving diagnosis across hundreds of
In multi-turn reinforcement learning, your custom reward function decides what the model actually learns. This post shows how to design a composite multi-turn reward for Amazon Nova Forge, execute model-generated code safely inside it, and instrument each component to catch the pitfalls that quietly collapse a reward.
366 points, 31 comments
Learn how to combine OpenAI-compatible endpoints on Amazon SageMaker AI with Amazon Bedrock AgentCore runtime to build a multi-agent workflow where each specialized agent uses the model best suited to its job. This post also shows how to get token-level observability from SageMaker endpoints that Strands Agents does not instrument by default.
498 points, 290 comments
Cursor is now a part of SpaceX
Autoformalization is commonly framed as translating natural-language mathematical statements into machine-verifiable formal languages such as Lean 4. However, faithful formalization requires more than translation. Models must map mathematical concepts to the complex hierarchy of types and definitions in formal libraries such as Mathlib, while ensuring that generated statements preserve the meaning
Laser frequency chirp is a ubiquitous dynamical process in semiconductor lasers, vital for frequency-modulated photonic systems. In the mid-infrared (MIR) and terahertz (THz) ranges, quantum cascade lasers (QCLs) are ideal sources with high power, narrow linewidth and compact size. While chirp dynamics in MIR QCLs have been studied, the transient chirp behavior of THz QCLs--particularly the therma
1170 points, 582 comments
Action-conditioned video world models require low-latency causal generation and reliable responses to game-native controls. Although causal distillation enables one- or few-step video synthesis, extending it to interactive world models remains challenging, as discrete keyboard states and continuous mouse motion must remain aligned with temporally compressed latent chunks during causal training and
We're excited to share that we are opening an office in Munich to serve companies across Germany, Austria and Switzerland.
Official Sierra technical update and publication covering Corporate.
Official Harvey AI technical update and publication covering Training Frontier Review Table Models With Applied Compute.
Official technical announcement and publication from Hugging Face covering State of Open Models: Summer 2026 Observations.
Official Harvey AI technical update and publication covering The Hidden Constraint in Contract Redlining Software.
Official Harvey AI technical update and publication covering The Judgment Contract Redlining Still Requires.
161 points, 49 comments
Official Harvey AI release and benchmark update covering Analyze the Full Scope of Legal Evidence, Not Just the Documents.
Official Anthropic technical update and publication covering Aug 14, 2026 Announcements How Claudeโs text watermark works.
We present a pulse-shaper-based dispersion-scan (d-scan) framework for the combined generation and characterization of polarization-tailored femtosecond laser fields. By integrating a programmable $4f$ pulse shaper with polarization-resolved d-scan measurements, the framework enables the programmable synthesis and reconstruction of complex time-dependent polarization states. We demonstrate its cap
AI co-scientists that generate hypotheses, retrieve related work, design experiments, execute code, and draft full papers are beginning to change how research is carried out. Despite this rapid progress, state-of-the-art systems remain researcher-agnostic: given a research goal, they optimize novelty, validity, or reviewer score while ignoring the individual scientist who will use the output. This
High-quality creative writing data for large language models (LLMs) remains dominated by story-centric data, limiting models' ability to follow the structural and functional conventions of diverse creative formats. We propose an attribute-guided genre expansion framework for scaling creative writing data beyond story generation. By separating thematic breadth from genre-form control, our framework
Popular facts are memorised more deeply during pretraining and resist removal longer than rare ones, yet existing LLM unlearning methods apply uniform gradient pressure regardless of training-data frequency. We propose the AdaPop (Adaptive Popularity) method, which combines local token confidence with a per-fact popularity-dependent exponent derived from an external proxy (e.g., Wikidata sitelinks
Open-weight language models are fine-tuned, quantized, pruned, and merged, yet their provenance is often undocumented. We study data-free white-box lineage verification: can weights alone reveal whether two compatible model checkpoints share ancestry? Residual training produces a shared identity-aligned component in branch products, so this structure alone cannot establish ancestry. We remove it a
Electrocardiogram (ECG) recordings are sensitive biomedical data, limiting the ability of hospitals and wearable devices to share raw signals for centralized model training. Federated learning addresses this practical privacy constraint by enabling collaborative model training while keeping raw biosignal data at their respective sources. However, federated ECG classification remains challenging du
Skills have emerged as a practical and effective approach for enhancing LLM agents at inference time through structured packages of knowledge. However, existing evaluations largely measure whether skills improve aggregated task success, leaving a more fundamental question underexplored: \textbf{When do skills help, why do they work, and where do they fail?} Through controlled experiments across va
Official CMU (Carnegie Mellon AI) technical update and publication covering SCS Alum Damion Shelton Named Associate VP and Executive Director of Swartz Center for Entrepreneurship.
Firetiger joins Cursor
712 points, 279 comments
968 points, 495 comments
Official technical announcement and publication from Hugging Face covering Record, train, and deploy from one place with Strands Agents, LeRobot, and Hugging Face Storage Buckets.
411 points, 166 comments
Official technical announcement and publication from Google DeepMind covering Introducing Gemini 3.7 Flash.
<div class="block-paragraph_advanced"><p><span style="vertical-align: baseline;">When enterprises transition from using simple chat assistants to autonomous, agentic workloads, they quickly run into a hard truth: Agents are prone to inaccurate insights when working with directly raw tables. </span></p> <p><a href="https://docs.cloud.google.com/bigquery/docs/graph-measures"><strong style="text-deco
At Sierra, we build agents using goals and guardrails so they can think and reason independently, whether originating a mortgage, disputing a charge, or helping select the right pair of skis. That freedom is what makes them so powerful โ and the guardrails they are given so important.
Two distinct Einstein--Hilbert sectors are related by a covariant correspondence bridge that maps stress-energy through bitensor kernels. The Bianchi identities enforce conservation, with bridge stresses balancing cross-sector exchange. For positive gravitational couplings and paired Minkowski backgrounds, a fixed metric-independent bridge leaves the quadratic action block diagonal, so the vacuum
220 points, 95 comments
Official CMU (Carnegie Mellon AI) technical update and publication covering Koedinger Wins Lifetime Achievement Award.
744 points, 310 comments
130 points, 185 comments
Cursor earns AIUC-1 certification for agent security and reliability
Cloud agents start 3x faster with builds
Official technical announcement and publication from Oxford (AIDC) covering Postdoctoral Research Associate on Integrated Approach to Computational Complexity.
Learn how startups use GPT-5.6 to build faster, more cost-efficient AI agents with smarter model selection and new Responses API capabilities.
The problem of computing \emph{diverse} solutions has recently emerged as an important area of study, motivated by applications in fairness, robustness, and security. Instead of returning a single feasible or optimal solution, the goal is to output a \emph{collection} of meaningfully different solutions, often measured by symmetric differences. Diverse variants have been studied using sparsificati
Preview Ultrafast, a new OpenAI API service tier that runs GPT-5.6 Sol up to 14ร faster. Powered by Cerebras, it delivers up to 750 output tokens per second.
OpenAI appoints Dali Rajic as Chief Revenue Officer to lead its global revenue organization and help businesses realize the full value of AI.
120 points, 89 comments
Official technical announcement and publication from Hugging Face covering What We Learned by Reproducing 2,200 papers from ICML.
Official Artificial Analysis technical update and publication covering Gemini 3.7 Flash: On the Intelligence vs. Time per Task Pareto frontier.
Official Artificial Analysis technical update and publication covering Announcing Optima: create a custom benchmark for your use case.
Official Harvey AI release and benchmark update covering The Brief: August 2026.
August 13, 2026
Agent Skills are today either hand-authored or produced in a single LLM generation pass, and consequently possess no closed loop through which they might improve from the interaction failures they actually cause. Recent work does close this loop, but derives its feedback from single-turn question-answering evaluation. The consequence is a sharp asymmetry: once the first round has patched the gaps
Electrocardiography (ECG), photoplethysmography (PPG), and phonocardiography (PCG) provide complementary views of the same cardiac cycle, yet existing cardiac foundation models are trained for a single sensing modality, leaving the shared physiology across sensors unexploited. We introduce CardioState-JEPA, a cardiac foundation model to learn a single shared representation jointly across ECG, PPG,
Open-ended real-world interaction admits multiple valid behaviors: an agent may answer directly, ask for clarification, provide progress updates, or confirm before acting. This flexibility breaks a core assumption behind group-based RL: rollouts compared within a group are no longer guaranteed to be behaviorally comparable. As a result, reward-model preferences over interaction style can distort r
Should you replace your text-embedding pipeline with a large language model? We answer this with a controlled, cost-aware comparison of ten LLMs across six families and 26 embedding models (118M to 14B parameters) on 37 tasks spanning classification, semantic textual similarity (STS), clustering, pair classification, and retrieval. In aggregate the two paradigms are effectively tied: the best LLM
Latent video generation relies on autoencoders to define a compact space in which generative models operate. Although video autoencoder architectures have evolved substantially, their latent spaces are still optimized primarily for pixel-level reconstruction and provide limited high-level semantic organization. A reconstruction-optimal latent space, however, need not be well suited to generative m
LLM coding agents issue Bash commands through interfaces that may serialize, wrap, and reparse model output. Matched execution scores alone cannot distinguish command-generation errors from failures introduced after generation. QuoteBench measures this boundary with exact final-state validation on 56 one-shot tasks from 14 incident-derived families, crossing the generation contract with the execut
6542 articles sourced historically ยท 100 per page