Artificial Intelligence
Artificial Intelligence news — Wednesday 18 November 2026
150 stories from 8 sources, every one in English. Each links to its publisher.
-
Nvidia, the first employee meeting "Go ahead"...
Nvidia emphasized the transition to a new management system in the first open employee meeting after the retirement of the former CEO of Jenson Huang. Some employees request the return of the sulfur, the executive promised to re-confirm…
-
[Nov.18] American Core System Lines...Nov., but AI Attack Horror
In the U.S., the main system of air control and banks, Microsoft (MS), Apple, and the National Security Agency (NSA), has a lot of anxiety about technology. There is no evidence that cyber attacks or AI causes, but recently, AI’s software…
- Introducing Gemini 4 Neon
- DaisyRose-27B: Pushing the Frontier of Post-Training Efficiency
-
Mathematical proof, powered by AI. And no hallucinations!
Every day brings new evidence of the awesome power of AI. There have been dramatic developments in mathematics: AI models are solving tough problems from Mathematical Olympiads and even some big open research problems. But AI agents are…
- Why your fine-tuning job takes longer than the FLOPs say
- Genie 4 is now available
-
How AI Agents Become a Tool for Attackers, and How to Stop It
Managing Security When You’re Deploying AI, Part 1: Compromised credentials, salaries inferred from budgets, and doctored documents show why agents need permission checks before the model ever sees the data. The post How AI Agents Become a…
- Time-Series Models in the Streamhouse: A New Frontier and Use Cases
-
Neurosymbolic Agents Explained, With a Z3 Example in Python
Starting from Air Canada's chatbot case, a tour of Logic-LM, Agent-C, and Amazon's checks shows the solver is sound but the English-to-logic step is not. The post Neurosymbolic Agents Explained, With a Z3 Example in Python appeared first…
-
The AI Hype Index: AI is everywhere
Like a nasty rash, AI seems to spread everywhere. Google crams AI Overviews down your throat even though it costs the company billions of dollars in lost ad revenue. Leave your apartment, and you might pass two people heaving an AI-powered…
-
When Qini Rewards the Wrong Uplift Model, Use Doubly Robust Scores
On 1 million simulated customers, calibrated Qini scored a Linear DR learner 0.0856 against 0.0826 for the true-CATE ranking that should have won. The post When Qini Rewards the Wrong Uplift Model, Use Doubly Robust Scores appeared first…
-
The Download: AI learns to smell and hydropower struggles
This is today’s edition of The Download , our weekday newsletter that provides a daily dose of what’s going on in the world of technology. This smell startup says its AI nose is as good as a human’s In a New Jersey lab, researchers are…
-
How to Estimate Dose-Response Curves When DoubleML Falls Short
Simulations test DoubleML against Kennedy's doubly robust estimator, with partial pooling across 200 stores and a DiD fix for hidden confounders. The post How to Estimate Dose-Response Curves When DoubleML Falls Short appeared first on…
-
OpenAI "Saver Security, Will Be Business Than Coding"...Producing New Growth Power
OpenAI is a big profit business than coding cybersecurity. We are already raising significant revenues, and we have strengthened our sales organization and partnerships with governments and businesses, and stood out on the market…
-
The main model such as Antropicy, OpenAI, and the ‘Game’ that belongs to the civilian and identified the military target
The model of major AI companies such as Antropice and OpenAI, Google, and DeepCike appeared to identify military targets to civilians and schools only by simple prompt operation. If you explain the war situation 'game', you can also check…
-
This smell startup says its AI nose is as good as a human’s
In a lab near Newark Airport in New Jersey, a team of around 30 experienced human sniffers are slowly working their way through every odor on Earth. Their careful notes—“nutty,” “peachy,” “marine”—are cross-referenced with a chemical…
-
Launch of 16 Perplex City, AI Agents to collaborate and validate
We are looking for a wide range of AI agents to deliver and validate the results of each other. We aim to increase the reliability of complex and accuracy critical work, such as financial analysis, corporate real estate, research, etc., in…
-
Google, Minai "Tsk Mode" is also available for free in Korea.
Google provides a free expansion to hundreds of millions in 6 countries including Korea and the United States. We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume…
-
AI·Robot·Autonomous driving...Opens 22 high-tech competitions in Gwangju
The 22 Colleges of Advanced Technology, such as AI model development, autonomous driving, and robot making, are held in Gwangju from 18 to 18. As an event to reveal the achievements of the joint education business participating in 67…
-
LIQUID AI Launches ‘D1.2’...AA Rating 1st
Requid AI has released the 'd1.2' and 'd1.2-ultra' of Omnimodal Crystal Models that handle text and images and audio. The top model d1.2-Ultra ranked No. 1 in the decision-making model evaluation of Atticians (AA). Requid AI announced that…
-
China AI Companies, Export Regulators, and Scale Model Cyber Attack Ability
China’s major AI companies reduce the attacking ability of models that can be exploited on cyber attacks such as hacking. While China government promotes the export restrictions of AI technology, which can attack core infrastructure and…
-
Hesaby’s AI New Drug Company, 6 months new 9.4 trillion investment attracted...
The AI New Drug Development Isomorphic Labs, founded by Dr. Hesavis Google Deep Mind Co-founder, attracted investments of $50 billion (about 6 trillion). In the last five months, we have a total of $1 billion (about $9 trillion) in six…
-
6600 billion investment in robot development like people... Genesis AI Corporate Value 4.6 trillion
The French startup GENESIS AI (Genesis AI), noted as the development of universal robots that moves hands and treats objects, was successful in attracting large investments. Since its foundation in the 2nd year, it has been recognized as…
-
Mondrian AI, a private AI cloud platform in Seoul.. . Original system Winback
The AI Cloud Professional Mondrian AI (CEO Hong) announced that it supplied ‘Runyour Cloud’ to the public AI cloud platform in Seoul National University. This is the ‘Win-back’ case, which was replaced by a third-party AI platform that was…
-
Jeonnam Gwangju, Sustainable AI Eyeglasses...132 Placement·Director Law Support
In the 2027th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year of the 19th year…
-
Green Soft, AI Smart Farm More AI in Cleanroom Technology
AI Solution Professional Green Soft (CEO Kim Jin-kuk) announced that it will lead to a cleanroom facility specialized in Korea’s eco-friendly company (CEO Kang Se-myung) and AI-based smart farm business. This affiliate will apply the AI…
-
30 states in the United States, "CounterDotAI" is required to strengthen the screening of each screening...
In the U.S., more than 30 state governments asked for a thorough investigation on the sales of AI chatbot services "Character.AI". In particular, a claim that the personal information and personal information of the children are included…
-
Nvidia expands global supply chain partners such as Samsung Electric and Ampoule
In response to increasing AI infrastructure demand, NVIDIA has strengthened the cooperation with supply chain partners such as Samsung Electric·Ampco Technology and expanding its substrate and advanced packaging production capacity. NVIDIA…
-
Inside the Codex sandbox
See how Codex runs commands on your machine while enforcing the filesystem and network limits you control across macOS, Linux, and Windows.
-
How Sumitomo Life’s leaders helped teams adopt ChatGPT Enterprise
Sumitomo Life uses ChatGPT Enterprise to research insurers, prepare proposals, and draft code for investment analysis. Early use by executives helped drive adoption.
-
Research acceleration: The jagged frontier
OpenAI measures how coding agents are accelerating AI research, where their performance remains uneven, and what this signals about recursive self-improvement.
-
27,000 people participated in the AI disaster safety diagnosis...
In the last year, we will continue to conduct civil safety education with AI-based disaster-related video based on disaster safety awareness based on mobile and QR code. This year’s program started at the end of September, with more than…
-
Nota, AI agent performance gains full stack optimization technology revealed..."Time 47%·Cover 60%"
Nata revealed ‘Full Stack Optimization’ technology that controls the workflow of AI agent to lighten AI models and GPU kernel optimization. The company side stated that the same AI model and GPU environment coding agent clad in the GPU…
-
OpenAI, NATO Cyber Security AI Supply... Expanding business in the European government and institutions
OpenAI supplies high-tech cyber security AI models to Northeast Ocean Pharmaceutical Organization (NATO) and expands government and international organizations. We are committed to protecting the core network of NATO and strengthening the…
-
5.65GW breakthrough in solar power generation in Mexico...More than 6 million installations
The capacity of distributed solar facilities in Mexico was beyond 5.65 GW of the year. The total dispersion-based power generation equipment has occupied 99.5% of the solar power system, and the power grid connection application processing…
-
Ember "Power supply cost 18% reduction in solar light, wind power and battery combination"
By adding wind power generation to solar and energy storage devices (ESS), the cost of meeting 90% of the annual power demands has been reduced from $105 per MWh (approximately USD 15 million) to $86 (approximately USD 12 million6000)…
-
GIST and AI University will lead the establishment of ‘AI Banddog Research Institute’.
The GIST, the Korea Institute of Science and Technology (GIST), will launch the AI university this year and promote the establishment of AI-based body research institutes. We are planning to establish a research system from semiconductor…
-
MS Introduces AI to perform high-end operations in ‘CopyLut’...
Microsoft (MS) introduced an AI agent ‘Autopilot’ that performs tasks without repeating instruction of users. Many employees and AI agents have added collaboration capabilities, and they also enhanced the system that manages integrations…
-
Managing frontier AI risk: From training to deployment
OpenAI explains how it manages frontier AI risk from training through deployment with model alignment, sandboxing, blocking safeguards, and monitoring.
-
Jenson Huang Daughter, MegaSpeed Engagement...
Jenson Huang’s daughter Madison Huang was known to be deeply involved in the support of the company’s foreign matters, including the Chinese Chip Miller’s hustle, Singapore’s Cloud Enterprise MegaSpeed and Envydia’s trade. As the…
-
Mi SEC, Jenson Sulfate After Envidia Survey...Identify whether or not to violate the investor
The U.S. Securities and Exchange Commission (SEC) was known to be a civil investigation on the Nvidia that meets the management crisis as the head of the former CEO of Jenson Huang. The Company uses ‘cookies’ that frequently save and…
-
Antropic, an IPO contest is an impressive review of $80 or more...
Antropic has been reviewing the implementation of higher interests of investors in front of the IPO. As a result of the change of AI, NVIDIA’s management team is expanding market uncertainties, but it is full of growth tax exempt from…
-
ProcReverser: Multi-Agent Black-Box Microarchitectural Reverse Engineering for Processor Vulnerability Analysis
arXiv:2611.20827v2 Announce Type: replace-cross Abstract: When processor implementation details are unavailable, black-box microarchitectural reverse engineering helps security researchers identify hardware conditions related to security…
-
LynnReal-Hand: Unlocking World Interaction with Hand Control
arXiv:2611.21479v2 Announce Type: replace-cross Abstract: Interactive video world models enable exploration through text and camera controls, yet direct object interaction requires finer control. Hands provide a natural interface, but…
-
What Matters for Cross-Modal Alignment
arXiv:2611.20000v2 Announce Type: replace-cross Abstract: The Platonic Representation Hypothesis posits that models trained on different modalities converge toward a shared representation. Related work aligns frozen image and text encoders…
-
Spec-Driven PBT: Deriving Semantic Invariants from Code Specifications for Agentic Property-based Testing
arXiv:2611.15920v2 Announce Type: replace-cross Abstract: Software is increasingly written by AI agents that produce an implementation from a natural language specification, a practice known as agentic coding. Code is now produced faster…
-
Rethinking "RL Generalizes, SFT Memorizes": The Role of SFT Data
arXiv:2611.14912v2 Announce Type: replace-cross Abstract: Large language models trained with Reinforcement Learning (RL) with verifiable rewards exhibit strong reasoning ability and broad generalization, whereas models trained with…
-
Attention at Absolute Zero
arXiv:2611.11937v2 Announce Type: replace-cross Abstract: Attention scores candidate supports of an answer, which we call witnesses, and reads the scores at a temperature. We build a higher-order attention layer whose state lives on pairs…
-
SkillPE: Creativity-Oriented Cinematic Skill Evolution for Text-to-Video Prompt Enhancement
arXiv:2609.34335v2 Announce Type: replace-cross Abstract: Achieving high-quality cinematic text-to-video generation remains challenging for non-experts, whose prompts often lack professional narrative and creative design. Existing prompt…
-
Structure Across Voices: Comparing acoustic-event type accumulation and sequence dependence across four vocal corpora using frozen audio encoders
arXiv:2609.16612v3 Announce Type: replace-cross Abstract: Comparing acoustic structure across very different vocal corpora is difficult because the recordings use different native event units, contain different amounts of sequence, and…
-
The Safeguard Worked. Is the LLM System Safer?
arXiv:2609.00519v2 Announce Type: replace-cross Abstract: Safeguards for large language model services are typically evaluated by their attack-blocking rates on test suites. Deployment decisions also depend on how much harmful assistance…
-
Sampling Luck Masquerades as Allocation Gain: Auditing Test-Time Budget Allocation for Neural Combinatorial Optimization
arXiv:2608.13087v2 Announce Type: replace-cross Abstract: Neural combinatorial optimization (NCO) solvers report the best of many sampled solutions per instance, and the sample count is, by convention, identical for every instance. Whether…
-
Semantic Prism: Generative Segmentation with an Explicit Semantic-Image Interface
arXiv:2608.11537v2 Announce Type: replace-cross Abstract: Conditional image generators can render segmentation maps as semantic RGB images, but color drift and boundary mixing can undermine direct decoding. We present Semantic Prism, a…
-
Role Steering of Language Models for Social Simulations
arXiv:2608.00023v3 Announce Type: replace-cross Abstract: Social simulations built from language-model agents need role-conditioned behavior that can be checked before agents are placed into a simulated population. We introduce an…
-
LLM Misuse Safeguards Rely on Untrusted Input: A Widespread, High-Risk Vulnerability
arXiv:2607.27951v2 Announce Type: replace-cross Abstract: Large language models now help professionals in fields such as cybersecurity and biology, yet the same answers would help an attacker. Deployers rely on misuse safeguards to keep…
-
Where Is the Cost of Third-Party API Routers in Agentic Software Development?
arXiv:2607.23624v4 Announce Type: replace-cross Abstract: Third-party API routers have become a common layer that unifies access across increasingly diverse LLM providers. In coding-agent workflows, high-autonomy operation is widely…
-
OpenMHC: Accelerating the Science of Wearable Foundation Models
arXiv:2607.16235v4 Announce Type: replace-cross Abstract: Mobile and wearable devices offer an unprecedented opportunity for continuous, passive health monitoring and active health coaching. However, the largest wearable datasets are…
-
Adapting Generalist Vehicle Models for High-Speed MPC Across Terrains
arXiv:2607.13319v5 Announce Type: replace-cross Abstract: High-speed off-road autonomy requires precise closed-loop control for a target vehicle while remaining robust across changing terrains. Recent forward kinodynamic (FKD) prediction…
-
Real-time fall detection based on vision for low-power edge platforms
arXiv:2607.12909v2 Announce Type: replace-cross Abstract: Falling detection is vital for elderly care and intelligent surveillance; however, prevailing vision-based approaches predominantly frame it as static pose classification or…
-
Spatiotemporal and Multi-Robot Motion Planning on Graphs of Space-Time Convex Sets
arXiv:2607.00444v2 Announce Type: replace-cross Abstract: Decoupled multi-robot motion planning avoids the high-dimensional joint configuration space but requires repeated single-robot planning among moving obstacles, where route choice…
-
Offline Preference-Based Trajectory Evaluation
arXiv:2606.17541v2 Announce Type: replace-cross Abstract: Offline evaluation of agentic systems often collapses trajectories into a binary terminal success value, discarding information about partial progress. This, in turn, results in…
-
Smooth Inliers, Erratic Outliers: Frequency-Aware Mamba for Two-View Correspondence Learning
arXiv:2606.04493v2 Announce Type: replace-cross Abstract: When shown matches between two photographs of the same cathedral, people can tell at a glance which ones are wrong: a window matched to its look-alike jumps against its neighbors…
-
Latent Q-Barrier Shielding for Safe In-Context Reinforcement Learning
arXiv:2605.25267v2 Announce Type: replace-cross Abstract: Safe in-context reinforcement learning (ICRL) adapts online from interaction history without test-time parameter updates while controlling episode cost under a safety budget. Under…
-
PEARL: Unbiased Percentile Estimation via Contrastive Learning for Industrial-Scale Livestream Recommendation
arXiv:2605.21752v2 Announce Type: replace-cross Abstract: Recommender systems trained on user interaction data are susceptible to behavioral intensity imbalance--a systematic distortion arising from heterogeneous engagement patterns across…
-
A closed-loop framework for perovskite precursor additive discovery
arXiv:2605.20242v2 Announce Type: replace-cross Abstract: Efficient discovery of precursor additives is essential for improving the performance of perovskite solar cells, yet the large chemical space makes conventional trial-and-error…
-
Efficient Generative Retrieval for E-commerce Search with Semantic Cluster IDs and Expert-Guided RL
arXiv:2605.14434v2 Announce Type: replace-cross Abstract: Generative retrieval offers a promising alternative by unifying the fragmented multi-stage retrieval process into a single end-to-end model. However, its practical adoption in…
-
AgentLens: Revealing The Lucky Pass Problem in SWE-Agent Evaluation
arXiv:2605.12925v5 Announce Type: replace-cross Abstract: Evaluation of software engineering (SWE) agents is dominated by a binary signal: whether the final patch passes the tests. This outcome-only view treats a principled solution and a…
-
Skypart : Weather-Robust Cross-View Geo-Localization via Prototype-Based Semantic Part Discovery
arXiv:2605.11654v3 Announce Type: replace-cross Abstract: The low-altitude economy is growing rapidly, increasing demand for reliable drone navigation. Common tasks such as delivery, inspection, and search-and-rescue rely on Global…
-
CyBiasBench: Benchmarking Bias in LLM Agents for Cyber-Attack Scenarios
arXiv:2605.07830v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly deployed as autonomous agents in offensive cybersecurity. In this paper, we reveal an interesting phenomenon: different agents exhibit…
-
WHBench: Evaluating Frontier LLMs with Expert-in-the-Loop Validation on Women's Health Topics
arXiv:2604.00024v3 Announce Type: replace-cross Abstract: Large language models are increasingly used for medical guidance, but women's health remains under-evaluated in benchmark design. We present the Women's Health Benchmark (WHBench)…
-
Human-Data Interaction, Exploration, and Visualization Systems in the AI Era: Challenges and Opportunities
arXiv:2603.05542v3 Announce Type: replace-cross Abstract: The rapid advancement of AI is transforming human-centered systems, with profound implications for human-AI interaction, human-data interaction, and visual analytics. In the AI era…
-
The Compute ICE-AGE: OPAL ONE, Continuity That Persists While Reasoning Is Idle
arXiv:2602.16736v4 Announce Type: replace-cross Abstract: The Compute ICE-AGE separates persistent application memory from probabilistic reasoning. We present OPAL, a CPU-resident C++17 graph engine that preserves record identities…
-
Intra-Agent Motion and Inter-Agent Relation Learning for Map-Free Trajectory Prediction
arXiv:2512.02368v3 Announce Type: replace-cross Abstract: Trajectory prediction is essential for safe planning and decision-making in autonomous driving. Map-free prediction improves deployment flexibility by removing the dependence on…
-
From Strategy Search to Policy Learning: MCTS-Guided Adversarial Co-Evolution for LLM Safety Alignment
arXiv:2511.19218v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) increasingly power web-facing applications, including conversational assistants and content services, where open user interaction exposes them to…
-
Cross-Epoch Bias in SGD-based Leave-One-Out Influence: Analysis and Accumulative Correction
arXiv:2510.26185v2 Announce Type: replace-cross Abstract: Training-data influence estimates support data selection and cleansing by predicting the effects of removing individual examples. The standard multi-epoch SGD-Influence Estimator…
-
TAR: Temporal Anchor-Constrained Reasoning for Video Temporal Grounding
arXiv:2508.07683v3 Announce Type: replace-cross Abstract: Video Temporal Grounding (VTG) aims to localize specific video segments corresponding to natural language queries. While recent Large Vision-Language Models (LVLMs) employ…
-
QUTCC: Quantile Uncertainty Training and Conformal Calibration for Imaging Inverse Problems
arXiv:2507.14760v3 Announce Type: replace-cross Abstract: While deep learning offers tremendous promise for scientific and medical imaging, any failures and hallucinations (predictions that do not coincide with reality) are hard to…
-
Recursive Learning-Based Virtual Buffering for Analytical Global Placement
arXiv:2506.17247v3 Announce Type: replace-cross Abstract: Due to the skewed scaling of interconnect versus cell delay in modern technology nodes, placement with buffer porosity (i.e., cell density) awareness is essential for timing closure…
-
Evaluating and Mitigating Skin Color Bias in Skin Lesion Classificatio
arXiv:2506.11627v2 Announce Type: replace-cross Abstract: Skin lesion classifiers may overlook performance disparity because of subtle skin tone variations by treating skin tone as a categorical attribute. This study proposes a method that…
-
Sage S2.1 Flash Technical Report
arXiv:2611.21072v2 Announce Type: replace Abstract: We introduce Sage S2.1 Flash, a 320-billion-parameter general-purpose mixture-of-experts model designed to reason over scientific evidence and act in interactive environments. Sage…
-
Information Limits of Replay-Free Recurrent State Migration
arXiv:2611.19345v2 Announce Type: replace Abstract: When a recurrent model is replaced, its old state may not determine what the new model would compute from the same history. We characterize the information that must be retained before…
-
Zero-Shot Weight-Space Programming for Prediction and Classification of Target Variables in Structured Datasets
arXiv:2611.19292v2 Announce Type: replace Abstract: Weight-space programming fixes every parameter of a feed-forward network by hand, with no gradient descent anywhere in the construction. Every demonstration so far starts from a target…
-
Lightfoot: Graph Retrieval-Augmented Generation with Exact, Model-Free Document Deletion
arXiv:2611.11010v2 Announce Type: replace Abstract: Graph-based retrieval-augmented generation (GraphRAG) improves multi-hop question answering, but deployed systems pay costs that research comparisons rarely measure: model calls at query…
-
Skill erosion, individual memory and the reversibility of AI adoption
arXiv:2611.04886v2 Announce Type: replace Abstract: Generative AI can improve immediate performance but may erode the human skills needed for future work. We consider a population of agents who perform a task either with their own skill or…
-
MatOS: An Agent Operating System for Materials Research
arXiv:2610.34203v2 Announce Type: replace Abstract: Materials research relies on scattered and growing data, knowledge, and simulation tools, and a single study often spans multi-day workflows across literature analysis, computation, and…
-
Same Simulation, Different Evidence: Where Retrieved Tool Output Costs LLM Agents Accuracy
arXiv:2610.26836v2 Announce Type: replace Abstract: LLM agents that compute the right answer can still answer wrongly. Agent harnesses increasingly keep long tool outputs out of the context window, saving them to files that the agent…
-
Screening LLM Judges Without Ground Truth: A Preliminary Evaluation of an AI Dental Hygiene Coach
arXiv:2610.24335v2 Announce Type: replace Abstract: Evaluating a fine-tuned coaching model with expert graders is slow and expensive, and an LLM judge cannot be trusted without ground truth. We present a staged pipeline that screens…
-
PhysAlign: A Benchmark for Evidence-Grounded Role Alignment in Multimodal Physics Reasoning
arXiv:2609.33319v2 Announce Type: replace Abstract: A key challenge in physics diagram understanding is correctly associating visual information with the physical entities, relations, and conditions it describes. Even when a value, symbol…
-
Multi-Granularity Context-Enhanced RAG over Multimodal Knowledge Graphs
arXiv:2608.25986v2 Announce Type: replace Abstract: Knowledge graph (KG)-based retrieval-augmented generation (RAG) leverages structured knowledge to provide large language models (LLMs) with high-quality external information. Building on…
-
Explore, Map, Remember, Decide: A Benchmark for VLM Readiness in Safety-Critical Scenarios
arXiv:2608.08077v2 Announce Type: replace Abstract: Vision-Language Models (VLMs) are often deployed as embodied controllers for robotics, yet it remains unclear how well they can explore an environment, maintain an internal map, and…
-
Test-Time Coverage: Test-Conditioned Data Curation for Deployment-Aware Learning
arXiv:2607.22697v2 Announce Type: replace Abstract: Deployed AI systems are often trained from broad candidate data pools, necessitating data curation towards the deployment test distribution. However, standard data curation methods score…
-
Optimization-like Reasoning in LLMs over Expanding Search Spaces
arXiv:2606.05464v2 Announce Type: replace Abstract: Verifiable reward training has improved mathematical and coding reasoning, but these domains capture only part of step-by-step decision making. Many real-world tasks require finding a…
-
CardioLoom: A Report-Guided Electrocardiogram Foundation Model for Cross-Domain Cardiovascular Phenotyping
arXiv:2605.25446v2 Announce Type: replace Abstract: Routine electrocardiogram (ECG) reports offer scalable supervision for cardiovascular representation learning. In this retrospective study, we pretrained CardioLoom, a report-guided ECG…
-
Scaling Causal Reasoning with Increasingly Complex Causal Simulators
arXiv:2605.09079v2 Announce Type: replace Abstract: Despite surpassing human performance across mathematics, coding, and other knowledge-intensive tasks, large language models (LLMs) continue to struggle with causal reasoning. A core…
-
Evaluating and Understanding Scheming Propensity in LLM Agents
arXiv:2603.01608v3 Announce Type: replace Abstract: As frontier language models are increasingly deployed as autonomous agents pursuing complex, long-term objectives, there is increased risk of scheming: agents covertly pursuing misaligned…
-
Do Multi-Agents Dream of Electric Screens? Achieving 100% on AndroidWorld Through Task Decomposition
arXiv:2602.07787v2 Announce Type: replace Abstract: We present Minitap's mobile-use, a multi-agent system for Android UI control, and evaluate it on the 116 tasks of AndroidWorld, on Android emulators in Minitap's cloud device…
-
CORE: Concept-Oriented Reinforcement for Bridging the Definition-Application Gap in Mathematical Reasoning
arXiv:2512.18857v4 Announce Type: replace Abstract: Large language models (LLMs) often solve challenging math exercises yet fail to apply the concept right when the problem requires genuine understanding. Popular Reinforcement Learning…
-
Toward AI epidemiology: a measurement standardization framework for prospective risk detection
arXiv:2512.15783v5 Announce Type: replace Abstract: This paper proposes a measurement standardization framework that compresses expert-AI interactions into structured, comparable fields for prospective risk detection in deployed AI…
-
Building verifiable databases at scale from the chemical literature via a multi-agent system
arXiv:2507.20230v4 Announce Type: replace Abstract: Accelerating chemical research with artificial intelligence requires high-quality databases. However, many reactions exist only as graphics drawn for human readers, which machines must…
-
Exploring Explainable Multi-agent MCTS-minimax Hybrids in Board Game Using Process Mining
arXiv:2503.23326v4 Announce Type: replace Abstract: Monte-Carlo Tree Search (MCTS) is a family of sampling-based search algorithms widely used for online planning in sequential decision-making domains and at the heart of many recent…
-
NS-Gym: A Comprehensive and Open-Source Simulation Framework for Non-Stationary Markov Decision Processes
arXiv:2501.09646v2 Announce Type: replace Abstract: Many real-world applications require decision-making where the environmental dynamics evolve over time. These non-stationary environments pose significant challenges to traditional…
-
AI Simulations Grounded in Rich Self-Reports Increase Accuracy and Reduce Distortion
arXiv:2411.10109v4 Announce Type: replace Abstract: AI simulations of people built from demographic prompts often misjudge individuals and distort sample-level patterns. We test whether rich self-report data reduce these problems. We…
-
StringQFT: A Multiuse Textbook Resource for AI-Assisted Theoretical Physics
arXiv:2611.23225v1 Announce Type: cross Abstract: We present StringQFT as a textbook collection for string constructions of quantum field theory and as infrastructure for AI-assisted theoretical physics. The collection comprises 18…
-
Do LLM Agents Still Need to Be Told How to Self-Improve?
arXiv:2611.23219v1 Announce Type: cross Abstract: LLM agents increasingly drive their own learning. Self-improvement methods delegate various components of model optimization to the models themselves. As models become more capable, this…
-
Reading Scene States: Uncovering Safety Failures in Vision-Language Models
arXiv:2611.23206v1 Announce Type: cross Abstract: Vision-language models (VLMs) can provide harmful assistance by interpreting visual events, even without an explicit request for the underlying harmful objective. We identify this…
-
Structure--Residual Inductive Biases for Language Model Pretraining
arXiv:2611.23197v1 Announce Type: cross Abstract: Next-token prediction is relied upon to learn both predictive behavior and internal representations. We ask whether representation structure itself should be an explicit inductive bias…
-
Retrieval-Augmented Correction for Generalist Medical Image Segmentation
arXiv:2611.23189v1 Announce Type: cross Abstract: Automated segmentation is routinely used as the first step in quantitative image-based clinical workflows, where predicted masks are subsequently reviewed and corrected by human operators…
-
Attribution Through the Decoder: Prediction-Space Explanations for Structured Outputs
arXiv:2611.23161v1 Announce Type: cross Abstract: Networks that predict angles, gaze directions, and rotations do not output these quantities directly: they emit raw coordinates that a fixed decoder, such as normalization, maps to the…
-
Mind the Gap: A Deployed Multi-Agent System for Documentation Maintenance
arXiv:2611.23142v1 Announce Type: cross Abstract: The "AI revolution" has enabled teams across the technology industry to deliver software features at unprecedented velocity. While code evolves faster than ever, customer-facing…
-
Natural Language Programming: Rethinking Language, Memory, and Cognition in the Era of Generative AI
arXiv:2611.23106v1 Announce Type: cross Abstract: Generative Artificial Intelligence is transforming the role of natural language in computational systems. In classical Natural Language Processing (NLP), language was treated as an object…
-
BugGen 2.0: Rethinking the Roles of LLMs for Scalable RTL Bug Synthesis
arXiv:2611.23091v1 Announce Type: cross Abstract: AI-assisted RTL localization and debugging need ground-truth bug datasets with known locations and observable failures, while testbench grading uses injected bugs to measure detection…
-
Runtime Verification of Multiple Natural Language Criteria for Agent Governance
arXiv:2611.23074v1 Announce Type: cross Abstract: AI agents are governed by hundreds of natural language criteria, yet existing verifiers cannot efficiently evaluate an agent's output against a large set of such criteria and report which…
-
ReTouchGS: Ultralight Render-free Aesthetic Color Enhancement and Transfer for 3D Gaussian Splatting
arXiv:2611.23054v1 Announce Type: cross Abstract: 3D Gaussian Splatting (3DGS) enables photorealistic scene reconstruction. Existing 3DGS color transfer and interactive editing methods are guided by reference images and user-specified…
-
Check Your Blind Spots: Decision-Guided Perception by Probing the Planner with Hypothetical Objects
arXiv:2611.23030v1 Announce Type: cross Abstract: Not every perception error changes the planner's action. Yet adaptive-resolution perception chooses where to spend high-resolution computation from image cues, which can neither reliably…
-
How Web Research Agents Interpret User Priorities
arXiv:2611.22989v1 Announce Type: cross Abstract: Web research agents gather information online and write cited reports, including financial analysis. Users can state a priority to guide this research, but a priority may not say how to…
-
Large Language Models Have Not Yet Transformed Science
arXiv:2611.22975v1 Announce Type: cross Abstract: The emergence of large language models (LLMs) in late 2022 has raised questions about the future of science, but assessing their impact requires understanding how scientific practices were…
-
CodeReel: Accelerating Code-to-Video Agents with Exact Frame Reuse and Blocking-Aware Scheduling
arXiv:2611.22967v1 Announce Type: cross Abstract: Coding agents make videos by writing and refining programs in a build--inspect--edit loop. A finished film costs a median of $2.85 in model calls and 11 minutes, over a third of that time…
-
Hierarchical Diffusion Transformer for End-to-End Image Generation
arXiv:2611.22962v1 Announce Type: cross Abstract: Pixel-space diffusion learns visual representations and image generation jointly in a single end-to-end training stage, without a separately pretrained visual encoder. However, plain…
-
Projected Penalty-Gradient Tracking for Communication-Efficient Byzantine-Resilient Federated Learning
arXiv:2611.22929v1 Announce Type: cross Abstract: We investigate communication-efficient federated learning over heterogeneous datasets in the presence of Byzantine adversaries capable of uploading arbitrary information. Penalty-based…
-
Stopping Is Not the Limit: Understanding and Eliciting Agent Capability in Long-Horizon Vulnerability Exploitation
arXiv:2611.22888v1 Announce Type: cross Abstract: Vulnerability exploitation tests whether language model agents can develop working exploits from known software flaws. Sequential test-time scaling can improve exploitation performance, but…
-
Learning Agent-Centric Vision-Language-Action World Models for Autonomous Driving
arXiv:2611.22880v1 Announce Type: cross Abstract: Recent progress in integrating vision-language-action (VLA) models with world modeling has significantly advanced end-to-end autonomous driving by allowing systems to reason about imagined…
-
WATERFALL: Workflow for Adaptive Training with Evolutionary Reward Formulation and Automated Learning Loops
arXiv:2611.22854v1 Announce Type: cross Abstract: Reward design remains a central bottleneck in reinforcement learning (RL), particularly in sparse-reward, long-horizon and partially observable settings requiring dependency chaining…
-
NexSight: Verifiable Tasks for Fine-Grained Agentic Long-Video Understanding
arXiv:2611.22790v1 Announce Type: cross Abstract: Fine-grained questions about long videos ask how many times an action occurs, in what order events unfold, or exactly when a state changes. Answering them can require finding brief moments…
-
PANO: Parallel Latent Space Neural Operator for Solving PDEs
arXiv:2611.22782v1 Announce Type: cross Abstract: Aggregating observations into a limited number of latent tokens and learning physical interactions in the latent space has emerged as an effective approach for Transformer-based neural…
-
ForceBody: Force-Paired Parametric Body Motion with Torque Uncertainty
arXiv:2611.22769v1 Announce Type: cross Abstract: Forces and torques drive human motion, central to biomechanics, robotics, and physics-aware animation, yet they are often absent from learning-based motion pipelines for body reconstruction…
-
Network-Exposed Configuration Attacks on LLM-Controlled Social Robots
arXiv:2611.22761v1 Announce Type: cross Abstract: Large language models (LLMs) in embodied robots create security dependencies beyond inference. We investigate a network-exposed configuration attack on the Pollen Robotics Reachy Mini Lite…
-
Stress Testing a Multimodal Large Language Model Grading Assistant on Kinematics Graph Reading
arXiv:2611.22726v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) can recognize graph features yet award credit without sufficient graphical support. This exploratory study examines selected grading outputs for…
-
Know When to Zoom: How Small Is Too Small for Vision-Language Models?
arXiv:2611.22709v1 Announce Type: cross Abstract: Online vision-language models often miss small details in large images, because the service shrinks the image, or rejects it, before the model ever sees it. A car 20 pixels wide in an 8000…
-
CULypso: A Clustered Federated Unlearning paradigm for Vehicular Ad-Hoc Networks
arXiv:2611.22682v1 Announce Type: cross Abstract: Unlearning has emerged as an essential security and privacy-preserving solution in Machine Learning, enabling networks to support their clients' rights for data erasure, as mandated by data…
-
FVScore: Specification-Driven Neuro-Symbolic Video Evaluation
arXiv:2611.22673v1 Announce Type: cross Abstract: We introduce Faithfulness Video Score (FVSCORE), a neuro-symbolic framework for specification-driven video generation evaluation. Existing evaluators rely on holistic MLLM judgments or…
-
AlignedHPO: Efficient Surrogate-Guided Loss Hyperparameter Optimization
arXiv:2611.22658v1 Announce Type: cross Abstract: Modern deep learning losses involve multiple hyperparameters, such as weights, scales, and margins, typically tuned manually or through expensive black-box optimization despite their…
-
Adaptive LLM-Driven Honeypots vs. LLM-Driven Attackers: Does Narrative Coherence Improve Deception?
arXiv:2611.22589v1 Announce Type: cross Abstract: Large Language Models (LLMs) enable honeypots to generate responses and deceptive artefacts on demand, yet existing approaches evaluate generated content one reply at a time and leave open…
-
Evolution Under Fire: Evolving Agent Safety Harnesses with Adaptive Injection Stressors
arXiv:2611.22583v1 Announce Type: cross Abstract: Tool-using language-model agents must act on information from untrusted emails, webpages, and documents. This creates a central security challenge: malicious instructions and legitimate…
-
Do LLMs Understand Network Traces? A Comparative Study of Reasoning over Raw and Filtered Traffic Data
arXiv:2611.22575v1 Announce Type: cross Abstract: The effective delivery and monitoring of ICT services rely heavily on network traffic analysis. However, interpreting complex packet captures remains a challenging task that requires…
-
JET: Just-Enough Exploration Tuning for Efficient Tool-Calling Agents in Repository-Level Question Answering
arXiv:2611.22539v1 Announce Type: cross Abstract: In repository-level question answering (QA), an agent explores a code repository with tools and should answer a question once enough information is gathered. Nevertheless, we find that LLM…
-
DGF-Asym: Adaptive Peer Weighting for Collective Estimation under Noisy and Faulty Sensing
arXiv:2611.22532v1 Announce Type: cross Abstract: Collective estimation under unreliable sensing requires robots to adaptively balance personal observations and social information from peers. We propose DGF-Asym, a decentralized collective…
-
Right Labels, Wrong Reasons: Signed Feature Attribution for Multi-Label Explainability
arXiv:2611.22515v1 Announce Type: cross Abstract: When a multi-label classifier assigns the right labels, it may do so for the wrong reasons. Diffuse feature combinations may vary across classifier families and lack a dominant feature…
-
Manufacturing Differential Sparsity: Planner-Hardware Co-Design for Efficient World Model Planning
arXiv:2611.22483v1 Announce Type: cross Abstract: JEPA-style latent world models let embodied agents plan toward new goals with model predictive control, but they are still too slow for real-life embodied control: sampling-based MPC…
-
LeViT-UNet++: Towards Balanced Fine-Grained Details and Context for Efficient Medical Image Segmentation
arXiv:2611.22482v1 Announce Type: cross Abstract: Efficient medical image segmentation is essential for clinical applications requiring fast and accurate delineation, such as surgical navigation, radiation therapy, and real-time…
-
MAVE: Reconstructing Editable 3D Blockouts from Film Shots
arXiv:2611.22464v1 Announce Type: cross Abstract: Animated 3D blockouts support previsualization and provide structural conditions for video generation, but manually reconstructing them from reference videos is costly. This limits both…
-
As Student Rollouts Continue, Visual Backdoors Keep Growing
arXiv:2611.22446v1 Announce Type: cross Abstract: Conventional knowledge distillation can facilitate the propagation of backdoors from teacher models to student models. However, how this risk manifests in On-Policy Distillation (OPD)…
-
Deal With It: On Better Schedules for Listwise Reranking
arXiv:2611.22392v1 Announce Type: cross Abstract: Listwise LLM rerankers cannot reliably rank hundreds of documents in a single call, and rely on schedules to orchestrate their calls. Parallel schedules split candidates into groups and…
-
RICoT: Recurrent Interleaved Chain-of-Thought for Page Stream Segmentation
arXiv:2611.22382v1 Announce Type: cross Abstract: Large language models (LLMs) have renewed interest in Page Stream Segmentation (PSS), splitting a multi-page file into its constituent documents. Yet published LLM methods still frame the…
-
FailureGen: Scalable Robot Failure Data Generation with World Models
arXiv:2611.22367v1 Announce Type: cross Abstract: Robot failures are difficult to collect at scale, limiting the range of failure scenarios represented in robot learning datasets. We present FailureGen, a method for generating synthetic…
-
An AI-Driven Assistant for Human-Centric Building Digital Twin Systems: Design, Implementation, and Comparative User Evaluation
arXiv:2611.22350v1 Announce Type: cross Abstract: Smart buildings generate streams of operational data, yet this richness rarely translates into accessible, decision-ready information for stakeholders who inhabit and manage them. Current…
-
Agent Trajectories as Evidence for Software Quality Assurance
arXiv:2611.22323v1 Announce Type: cross Abstract: Autonomous software agents increasingly plan, implement, and test software with limited human oversight. Their decisions may introduce implicit assumptions that final code and passing tests…
-
Perception into Imagination: Executable Memory for World Models
arXiv:2611.22294v1 Announce Type: cross Abstract: Physical agents must use what they previously saw or touched after visibility or contact is lost. Their actions change the implications of these remembered facts. We introduce executable…
-
Neural Network Path Optimization for Finding Transition States on a Machine Learned Potential
arXiv:2611.22272v1 Announce Type: cross Abstract: Transition states govern chemical reaction rates and outcomes, yet locating them remains a central challenge in computational chemistry. Double-ended search methods find transition states…
-
Anticipating or Speculating? Making Futures Pay in AI Ethics
arXiv:2611.22268v1 Announce Type: cross Abstract: Philosophers have recently been trying to develop uniform decision rules for instances in which physicians' judgments diverge from the recommendations of medical artificial intelligence…