PyPI / crates.io / RubyGems / Go index / NuGet
frida 17.19.0
Dynamic instrumentation toolkit for developers, reverse-engineers, and security researchers
Science & Research
Stories tagged science & research.
Utility Dive / pv magazine / Energy-Storage.news
Solar panel waste silver turned into catalyst for carbon monoxide production
South Korean researchers have recovered high-purity silver from end-of-life silicon solar panels and used it to produce electrocatalysts for converting carbon dioxide into carbon monoxide. The recycled-silver electrode sustained CO production for more than 600 hours, with its rough surface helping…
Utility Dive / pv magazine / Energy-Storage.news
Inclusive leadership: turning diverse perspectives into innovation and impact
At Solar & Storage Live UK 2026, Women in Solar+ Europe (WiSEu) explored the leadership behaviours that help diverse teams contribute, challenge ideas and drive innovation. A Solar+ Leaders panel extended the conversation to the trust and responsibility needed to deliver lasting impact. The post…
Canary Media / Heatmap / Latitude Media
Permitting Reform Talks Trail Off
Current conditions: After starting to dissipate, Tropical Storm Fay has regenerated as a “zombie storm” in the Atlantic • Tropical Storm Nolo is barreling straight toward Hawaii’s Big Island, bringing yet more floods and powerful winds to a state still reeling from an unusual string of tempests •…
Utility Dive / pv magazine / Energy-Storage.news
The Hydrogen Stream: Germany, Austria and Luxembourg launch €2.12 billion program for sustainable aviation fuel
In other news, African green hydrogen projects secure new funding and Mauritania advances the Aman project. Moreover, Ansaldo Green Tech is expands electrolyzer manufacturing, as Toyota, Iveco and Hyster advances hydrogen-powered heavy-duty transport. The post The Hydrogen Stream: Germany, Austria…
Utility Dive / pv magazine / Energy-Storage.news
India adds 50.6 GW of solar module and 9.7 GW of cell manufacturing capacity in 1H
As of June 2026, India’s cumulative solar module manufacturing capacity stood at 261.7 GW, while cell manufacturing capacity reached 36.6 GW. The post India adds 50.6 GW of solar module and 9.7 GW of cell manufacturing capacity in 1H appeared first on pv magazine Global.
Utility Dive / pv magazine / Energy-Storage.news
Weaker monsoon lifts irradiance in western India as August storms dim the east
In a new weekly update for pv magazine, Solcast, a DNV company, reports that northwestern and southern India and Pakistan recorded above-average solar irradiance during the 2026 summer monsoon, benefiting major PV-producing regions, while tropical depressions drove irradiance sharply below average…
Utility Dive / pv magazine / Energy-Storage.news
US ROUNDUP: BESS portfolio deals, investment and partnerships for Energy Vault, SMT Energy, NeoVolta and SK On
Recently, US BESS companies Energy Vault and Goshe, SMT Energy and Climate Adaptive Infrastructure, and NeoVolta and SK On have made major deals and secured strategic partnerships.
Utility Dive / pv magazine / Energy-Storage.news
China and U.S. module prices hold ahead of Golden Week and Section 232 price floors
In a new weekly update for pv magazine, OPIS, a Dow Jones company, provides a quick look at the main price trends in the global PV industry. The post China and U.S. module prices hold ahead of Golden Week and Section 232 price floors appeared first on pv magazine Global.
SecurityWeek
Windows, Linux, Android File Notification Systems Leak User Activity
Researchers show that file-change notification systems can leak keystroke timing, browsing activity, and WhatsApp media events. The post Windows, Linux, Android File Notification Systems Leak User Activity appeared first on SecurityWeek.
Utility Dive / pv magazine / Energy-Storage.news
US power conversion manufacturers in spotlight: Power Electronics builds Houston factory, Heron Power lands RWE pilot
Spain-headquartered Power Electronics has opened a new factory in Houston, Texas, as the US market increasingly looks to build domestic content supply chains.
Utility Dive / pv magazine / Energy-Storage.news
Spain awards €265 million for 49.9 GWh of pumped hydro storage
Pumped hydroelectric energy storage is advancing in Spain, with 11 projects backed by €265 million in public subsidies from a government program. The post Spain awards €265 million for 49.9 GWh of pumped hydro storage appeared first on pv magazine Global.
The Quantum Insider / Quantum Computing Report
Mphasis and Copa Airlines Host Quantum Computing Challenge for Airline Optimization
Insider Brief PRESS RELEASE — Mphasis, (BSE: 526299; NSE: MPHASIS), a global AI-led, platform-driven technology solutions provider, in partnership with Copa Airlines, recently hosted a Quantum Computing Challenge, an international innovation contest. The challenge brought together teams from…
Utility Dive / pv magazine / Energy-Storage.news
German developer tests roe deer passages at solar park
German solar developer Südwerk has integrated roe deer passages into the perimeter fencing of its 20 MW Memmelsdorf solar park in Bavaria to preserve existing wildlife corridors. The company plans to consider similar measures in future projects following the experience gained at the site. The post…
Utility Dive / pv magazine / Energy-Storage.news
Decommissioning to drive 23% of solar buildout in the 2040s, says Wood Mackenzie
Wood Mackenzie’s latest report says the need to replace ageing infrastructure will bring a more than 60% increase in equipment sales in 2050 when compared to today's levels. The post Decommissioning to drive 23% of solar buildout in the 2040s, says Wood Mackenzie appeared first on pv magazine…
Help Net Security
MacSync info-stealing malware hides malicious commands in an iCloud calendar
A new MacSync variant targets Mac users with an infostealer and persistent backdoor designed to steal credentials, crypto wallet data, and files, according to Kaspersky. Researchers found the malware spreading through a crypto wallet app called Toria, which had its own website and was promoted on X…
Fierce Biotech / Fierce Healthcare / MedCity / GEN
New CRISPR-Based Tool Enables Genome-Wide Mutagenesis of Bacteriophages
A CRISPR-based method for disrupting genes across bacteriophage genomes has been developed, creating a faster route to uncover gene functions and engineer phages for potential therapeutic and biotechnology applications. The post New CRISPR-Based Tool Enables Genome-Wide Mutagenesis of…
Utility Dive / pv magazine / Energy-Storage.news
Denmark takes first steps toward marine floating PV
Denmark has approved a three-year research project to test three floating PV technologies totaling around 115 kW in Roskilde Fjord. The project will assess their technical performance, environmental impact and economic viability under Danish marine conditions. The post Denmark takes first steps…
The Quantum Insider / Quantum Computing Report
DigiCert Launches Quantum Central for Post-Quantum Cryptography Readiness
Insider Brief PRESS RELEASE — DigiCert, a global leader in intelligent trust, today announced the general availability of DigiCert Quantum Central, part of the DigiCert ONE platform, that helps organizations manage and demonstrate progress toward post-quantum cryptography (PQC) readiness. First…
The Quantum Insider / Quantum Computing Report
Azulene Labs Raises $3.4 Million Pre-Seed Round to Build Physics-Grounded AI Foundation Models for Chemistry
Azulene Labs, a molecular simulation developer, has raised $3.4 million in pre-seed funding. The company, founded by Dr. Nicolas Sawaya, aims to revolutionize chemical discovery by building physics-grounded AI foundation models. Their platform offers significantly improved computational efficiency…
Utility Dive / pv magazine / Energy-Storage.news
US team claims stability record for 16.19%-efficient 2D/3D tin perovskite solar cell
US researchers developed a 2D/3D lead-free tin perovskite solar cell using a chlorinated spacer, achieving 16.19% efficiency while improving crystallinity and resistance to oxygen and moisture. The device retained 95% of its initial efficiency after more than 1,600 hours in dry air and 80% after…
ANZ: iTnews / SmartCompany / Startup Daily
Abandoned QR code subdomains open to hijacking, researcher says
'QR Jacking' easy and fast way to exploit a vulnerability in a trusted ecosystem.
The Quantum Insider / Quantum Computing Report
Evaluating Quantum Generative Models on Real Satellite Radar
Insider Brief PRESS RELEASE — IonQ researchers applied a quantum generative machine learning model to help detect changes in highly complex satellite image data. The results, detailed in a recently published paper, reveal that using quantum-based analysis shows promise for detecting and predicting…
Utility Dive / pv magazine / Energy-Storage.news
The impact of phase change material on rear-side PV heat transfer
A year-long outdoor experiment in Milan found that a PCM-backed PV module had 45.2% lower median rear-side convective heat flux than a reference module. The researchers also found that PCM’s thermal history affected heat rejection, even when its rear surface was at the same temperature. The post…
Canary Media / Heatmap / Latitude Media
How dirty is the power generated in your state?
See more from Canary Media’s “Chart of the Week” column . Have you ever gazed up at power lines and wondered how clean the electrons passing along them are? The answer is very different depending on where you are. Just look at the spectrum of blues in the map above. It charts the carbon intensity…
Utility Dive / pv magazine / Energy-Storage.news
Trane unveils cold-climate heat pump operating down to -29 C
Trane has launched a residential cold-climate air-source heat pump that operates at temperatures from -29 C to 46 C and is available in 3-ton, 4-ton and 5-ton versions using R-454B refrigerant. The system can deliver up to 107% of its rated heating capacity at -15 C and retain up to 85% at -29 C…
The Quantum Insider / Quantum Computing Report
UTM Becomes Base For Malaysia’s First Quantum Computing Hardware Company
Insider Brief PRESS RELEASE — Universiti Teknologi Malaysia (UTM) is strengthening its role in Malaysia’s emerging quantum technology ecosystem as AQSolotl Pte. Ltd., a Singapore-based quantum technology company, establishes its research, development and commercialisation (R&D&C) operations at UTM…
The Quantum Insider / Quantum Computing Report
DigiCert Announces General Availability of Quantum Central Platform for PQC Migration Management
DigiCert has launched Quantum Central, an enterprise management suite within its DigiCert ONE platform, to help organizations manage their transition to quantum-safe encryption. This tool consolidates cryptographic assets, tracks policy compliance, and automates remediation workflows, addressing…
The Quantum Insider / Quantum Computing Report
Qambria Secures $2.4M Pre-Seed Funding to Commercialize Sub-Microsecond Classical Control Infrastructure for Fault-Tolerant Quantum Systems
Qambria AG, a Swiss quantum infrastructure developer, has successfully closed a $2.4 million pre-seed funding round led by Syntropy. The company aims to commercialize a classical engine for quantum computing, focusing on software and control orchestration to integrate quantum processing units with…
Europe/UK: Tech.eu / Sifted / TNW / Silicon Republic / UKTN / Computer Weekly / heise / Golem / t3n / Maddyness / FrenchWeb
‘Physics is not simply a collection of equations’ for this UL researcher
Dr Maria Alhajji is involved in teaching, research, student supervision, and contributing to departmental education and outreach initiatives. Read more: ‘Physics is not simply a collection of equations’ for this UL researcher
The Quantum Insider / Quantum Computing Report
QuiX Quantum Partners with Tohoku Electronic Industrial for Exclusive Distribution in Japan
QuiX Quantum has partnered with Tohoku Electronic Industrial Co., Ltd. (TEI-C) for exclusive distribution of its integrated photonic processing hardware in Japan. This includes the Alquor 2.0 quantum photonic processor and the PACU control unit. The partnership aims to accelerate access to…
The Quantum Insider / Quantum Computing Report
Mesa Quantum Raises $11.8M to Scale Chip-Scale Quantum Sensors and Alternative PNT Infrastructure
Mesa Quantum has raised $11.8 million in funding, led by Playground Global, bringing their total venture funding to nearly $16 million. This capital will accelerate the development and manufacturing of their chip-scale quantum sensors and alternative PNT infrastructure. Their technology, which…
The Quantum Insider / Quantum Computing Report
Infleqtion Achieves 30 Entangled Logical Qubits on Sqale Neutral-Atom Quantum Processor
Infleqtion has achieved 30 entangled logical qubits on its commercial Sqale quantum computing platform, utilizing 80 physical neutral-atom qubits. This breakthrough, announced at Quantum World Congress 2026, was enabled by an AI-assisted logical circuit discovery and software-based loss correction…
The Quantum Insider / Quantum Computing Report
Mphasis and Copa Airlines Partner on Global Quantum Computing Challenge for Airline Disruption Management
Mphasis and Copa Airlines partnered on a global Quantum Computing Challenge, bringing together academic teams to develop quantum optimization solutions for airline passenger re-accommodation during disruptions. The hackathon leveraged quantum algorithms and real operational data to improve…
Utility Dive / pv magazine / Energy-Storage.news
Solar sleeves turn telecommunications towers into renewable power generators
Victorian telecommunications and electrical infrastructure services company Decon Corporation has finalised a trial of solar laminate panels wrapped on a National Broadband Network fixed wireless monopole. The post Solar sleeves turn telecommunications towers into renewable power generators…
The Quantum Insider / Quantum Computing Report
IonQ Demonstrates Quantum Generative Modeling Advantage for High-Resolution Satellite Radar Change Detection
IonQ's research shows that quantum generative machine learning models on their trapped-ion quantum processing units can more accurately detect ground-level changes from satellite radar imagery, especially with high-resolution data causing sparse pixel statistics. Their Quantum Circuit Born Machines…
The Quantum Insider / Quantum Computing Report
Florida International University Contracts IonQ for On-Premises Superion 256 Trapped-Ion Quantum Computer
IonQ has signed an agreement with Florida International University (FIU) to provide and install an on-premises IonQ Superion 256 trapped-ion quantum computer by late 2027. This makes FIU IonQ's main academic partner in Florida and the first to have a 256-qubit system in the state. The quantum…
The Quantum Insider / Quantum Computing Report
QC Design Unveils Meridian: Purpose-Built AI Architecture System Demonstrates Over 10× Reduction in Logical Error Rates
QC Design has introduced Meridian, an AI-powered platform for optimizing fault-tolerant quantum computing (FTQC) hardware. Meridian has demonstrated a 14.6x median reduction in logical error rates compared to existing methods and a 43% reduction against other frontier AI agents like GPT-6 Astra…
PyPI / crates.io / RubyGems / Go index / NuGet
superfermion 0.1.12
High-performance quantum computing framework with a Python API and Rust simulation core
The Quantum Insider / Quantum Computing Report
memQ Open-Sources Industry-First Modality-Agnostic Distributed Quantum Compiler (memQ DQC)
memQ Inc. has open-sourced its Distributed Quantum Compiler (memQ DQC), a modality-agnostic software framework for scaling quantum computing across multi-vendor QPU networks. This release aligns with U.S. federal directives and allows developers to model multi-QPU execution and analyze resource…
The Hacker News
Cloudflare Fixes Flaw That Let One Container Read Another Customer's Leftover Disk Data
A flaw in Cloudflare Containers let a paying customer read data that other customers' containers had left behind on the same server, Cloudflare and the researchers who found it said on Thursday. The data came from disk space that earlier containers had used and given up, not from any live workload…
The Quantum Insider / Quantum Computing Report
memQ Releases memQ DQC: An Open-Source Framework for Distributed Quantum Computing
memQ Inc. has released memQ DQC, an open-source Python framework for distributed quantum computing. It compiles and schedules quantum programs across multi-processor quantum architectures, addressing challenges in scaling quantum processing units beyond physical limitations. The framework allows…
Cointelegraph
Researchers propose Zcash-style private Bitcoin transfers without a soft fork
Shielded Bitcoin would use zero-knowledge proofs and encrypted notes to hide transaction details without changing Bitcoin consensus, but researchers have raised questions about anonymity and quantum resistance.
arXiv
Safety Nudges: User-Facing Interventions for Real-Time AI Risk Awareness
arXiv:2609.26865v2 Announce Type: replace-cross Abstract: Conversational AI systems can pose safety risks to their users such as hallucination, sycophancy, overconfidence, and anthropomorphism, but these risks are difficult for users to detect during everyday use. We introduce Safety Nudges, a…
arXiv
Calibration is the Bottleneck: An Action-Class Diagnostic of Multi-Turn Tool-Calling
arXiv:2609.00949v2 Announce Type: replace-cross Abstract: Multi-turn tool calling is a core evaluation scenario for large language model (LLM) agents. On public tool-calling benchmarks, open-weight models now approach or even surpass closed-source frontier models in aggregate accuracy. However…
arXiv
Hardware Keystores for AI Agent Signing Workflows: A Zero-Trust MCP Enforcement Architecture
arXiv:2608.06130v2 Announce Type: replace-cross Abstract: AI agents increasingly sign Git commits, certify documents, and attest release artifacts on behalf of their operators, using private keys that live in software-accessible locations (plaintext files, environment variables, container memory)…
arXiv
PAWS: Policy-driven Agentic World Simulation
arXiv:2609.28547v1 Announce Type: new Abstract: Policy interventions propagate through public communication, institutional decisions, and stakeholder responses, yet datasets for financial multi-agent simulation rarely connect these processes to temporally aligned historical evidence. We introduce…
arXiv
LAYERSCOPE: A Layerwise Characterization of Video and Multimodal Learned Representations
arXiv:2609.28086v2 Announce Type: replace-cross Abstract: We propose LAYERSCOPE, a label-free, layerwise framework that aims to characterize a model's learned representations in video and multimodal settings. Evaluating downstream performance using representations from final or intermediate layers…
arXiv
TacSushi: Tactile-Grounded World-Action Modeling for Dexterous Sushi Manipulation
arXiv:2609.19613v2 Announce Type: replace-cross Abstract: Dexterous food manipulation requires control under deformation, occlusion, and uncertain contact. We present TacSushi, a tactile-grounded, Cosmos3-based world-action policy that learns from recorded future consequences while acting on…
arXiv
How broad is that claim? Mapping Generalisation in NLP Research
arXiv:2609.14770v2 Announce Type: replace-cross Abstract: Generalisations are common in scientific communication, even though they are semantically ambiguous. An automated method is needed to identify and categorise claims according to their level of generalisation, in order help detect an…
arXiv
Evolve Vision-Language-Action Model into an Agent with On-the-fly Tool-use
arXiv:2608.14047v3 Announce Type: replace-cross Abstract: This paper integrates end-to-end Visual-Language-Action (VLA) models with agentic tool-use to propose Agentic Robot with Tool-use (ART). ART is a tool-injection framework that tunes any VLA model to leverage off-the-shelf tool modules for…
arXiv
Cryptographically verifiable authorization for autonomous AI agents: a falsifiable hypothesis and proof of concept
arXiv:2607.21325v3 Announce Type: replace-cross Abstract: Autonomous AI agents increasingly execute actions, invoke tools, and operate on protected resources with limited human oversight. Existing authentication and authorization mechanisms establish identity and delegate authority but do not…
arXiv
Topological Signatures of Cyber-Attack Classes in Natural Visibility Graph Representations of Network Traffic
arXiv:2609.26990v2 Announce Type: replace-cross Abstract: Natural Visibility Graph (NVG)-based representations provide a promising approach for capturing structural patterns in sequential network traffic. However, whether different cyber-attack classes exhibit distinctive topological signatures in…
arXiv
TERN: A Delta-rule Memory with a Seasonal Reference and Online Adaptation for Epidemic Forecasting
arXiv:2609.18407v2 Announce Type: replace-cross Abstract: Weekly influenza surveillance counts guide vaccine distribution and public-health alerts, yet they are hard to forecast. Each region offers only a few seasons, waves shift in timing and height every year, and information that helps while a…
arXiv
Aero Hand Open: A Simulation-Ready Tendon-Driven Hand for Dexterous Manipulation Learning
arXiv:2608.28578v2 Announce Type: replace-cross Abstract: Tendon-driven hands are anthropomorphic, and moving the actuators off the joints is what makes a hand of this capability affordable to build. Two effects produce that saving. Routing force through a cable removes the requirement that a motor…
arXiv
Trident : How to Break Deep Reinforcement Learning Cyber Defenses (Agentic)
arXiv:2608.04317v2 Announce Type: replace-cross Abstract: Autonomous cyber defense systems based on Deep Reinforcement Learning (DRL) have attracted significant research attention, yet remain evaluated almost exclusively against static, heuristic red agents, leaving their robustness against…
arXiv
Analyzing Defensive Misdirection Against Model-Guided Automated Attacks on Agentic AI Systems
arXiv:2606.20470v3 Announce Type: replace-cross Abstract: Agentic AI systems increasingly rely on language-model components to interpret instructions, process external data, invoke tools, and coordinate with other agents. These capabilities make prompt-injection and jailbreak attacks more…
arXiv
TWIST: A Proposed Benchmark for Intervention Quality in Conversational Memory, with a Human-Validated Draft-Alignment
arXiv:2609.28575v1 Announce Type: new Abstract: Long-conversation memory benchmarks increasingly test recall and prompted knowledge updates, and recent work studies evolving user beliefs and memory state. TWIST is a proposed benchmark suite for a complementary, unmeasured property: intervention…
arXiv
TW3Cast: A Frozen Router of Lightly Fine-Tuned Foundation Models for Time-Series Forecasting on GIFT-Eval, Selected Entirely on the Training Split
arXiv:2609.28506v1 Announce Type: new Abstract: TW3Cast is a time-series forecasting system that reaches position 3 of 130 entries on the GIFT-Eval benchmark by mean MASE rank, as of 2026-09-14. The two entries above it belong to the leaderboard's agentic category, multi-step systems that use…
arXiv
Broadening Uncertainty Estimation for Audio Question Answering Across Methods, Formats, and Inputs
arXiv:2609.28879v1 Announce Type: cross Abstract: Audio-language models can produce confident answers unsupported by the audio, motivating uncertainty estimates that identify unreliable responses. We compare probability-based, sampling-based, self-verification, evidential, and contrastive measures…
arXiv
Design and Evaluation of LLM Chaining-Based Task Planning for General Purpose Service Robots
arXiv:2609.29043v1 Announce Type: cross Abstract: General Purpose Service Robot (GPSR) tasks, as defined in the RoboCup@Home benchmark, require robots to interpret diverse natural language commands and generate multi-step action sequences in real home environments. Conventional Single Prompt (SP)…
arXiv
DAWN: Noise-Robust Quadruped Parkour via Depth-Denoising World Models
arXiv:2609.29092v1 Announce Type: cross Abstract: Vision-based legged locomotion methods assume clean depth at training time and rely on hand-tuned post-processing filters at deployment. However, filter parameters are rarely disclosed, hindering reproducibility, and performance degrades…
arXiv
Med-AR: Autoregressive Vision-Language Pretraining for Long-Tailed Chest X-Ray Classification and Uncertainty-Aware Evaluation
arXiv:2609.29156v1 Announce Type: cross Abstract: Long-tailed chest X-ray classification requires visual representations that capture both common abnormalities and subtle, infrequent findings. We propose Med-AR-8B and Med-AR-2B, two radiology-native autoregressive vision-language models pretrained…
arXiv
No More Free Lunch: Corpus Task Complexity Matters as Corpora Grow
arXiv:2609.29245v1 Announce Type: cross Abstract: Given a large corpus, the questions one might ask can vary -- from "When was the first human heart transplant?" to "What are all the contradictory claims in this literature?" -- but what makes some questions more challenging than others? In this…
arXiv
TP-CRIV: A Framework for Third-Party Challenge-Response Identity Verification of AI Models
arXiv:2609.29264v1 Announce Type: cross Abstract: Artificial intelligence (AI) models are increasingly deployed through remote services, making model misappropriation a growing concern. Existing approaches, including watermarking, fingerprinting, and model similarity analysis, primarily rely on…
arXiv
WST-Graph: Topology-Preserving Wavelet Scattering Front-End for Speech Deepfake Detection
arXiv:2609.29372v1 Announce Type: cross Abstract: The acoustic front-end determines which forensic cues a speech deepfake detector can exploit. The wavelet scattering transform (WST) provides stable multiscale coefficients with explicit coordinates, yet direct flattening obscures the parent…
arXiv
AgriCountDINO: Parameter-Efficient Exemplar-Guided Counting and Localization in Agriculture
arXiv:2609.29460v1 Announce Type: cross Abstract: Accurate counting and localization of plants and their organs support phenotyping and yield estimation, yet target appearance, scale, and density vary widely across species and imaging conditions. Exemplar boxes specify the target without…
arXiv
GeoRefer-Bench: A Benchmark from Referring Pixels to Verifiable Geospatial Reasoning
arXiv:2609.29541v1 Announce Type: cross Abstract: Referring segmentation in overhead imagery is inherently relational: a query may ask for the buildings north of the road or the pond closest to a residential area, so the correct referent can contain one object, several objects, or none. Existing…
arXiv
Evidence-Driven Differential Diagnosis of Malignant Melanoma
arXiv:2609.29613v1 Announce Type: cross Abstract: We present a modular and multi-level framework for the differential diagnosis of malignant melanoma. Our framework integrates contextual information and evidence at the lesion, patient, and population levels, enabling decision-making at each level…
arXiv
Graph, Loop, and Harness Engineering for Zero-Trust Agentic Data Engineering and Analytical Processing
arXiv:2609.29668v1 Announce Type: cross Abstract: Large language model agents increasingly automate data workflows, but end-to-end cloud data engineering and analytical execution require reliable coordination across code, data, infrastructure, and runtime environments. We present two zero-trust…
arXiv
Revalidation Beats Stateful Routing for Scientific Surrogates Under Distribution Shift
arXiv:2609.29715v1 Announce Type: cross Abstract: Surrogate models are often chosen during development and then left in place as new measurements arrive. That practice becomes risky when noise, input support, or physical parameters change. We asked whether such changes call for a stateful adaptive…
arXiv
Benchmarking and Domain Adaptation of Automatic Speech Recognition (ASR) for Adolescent Health Communication in Ghanaian Languages
arXiv:2609.29798v1 Announce Type: cross Abstract: This paper presents an end-to-end study of automatic speech recognition (ASR) for adolescent health communication in three Ghanaian languages (Twi, Dagbani, and Ewe). The work proceeds in three connected stages; First, we benchmark five ASR systems…
arXiv
Working with Agentic `Teammates': When a New Organizational Actor Collides with the Human Ecosystem of Work
arXiv:2609.29901v1 Announce Type: cross Abstract: Enterprise AI is transitioning from single-user, reactive tools toward proactive, multi-user 'teammates,' but our empirical understanding of this transition is limited. In this paper, we present an in-situ qualitative study of a persistent…
arXiv
Tracking States or Tracking Cosets? An Algebraic Account of Learned State Tracking
arXiv:2609.29951v1 Announce Type: cross Abstract: State tracking requires composing a sequence of updates, but accuracy alone does not reveal what a model has learned. We study neural networks trained to predict the running product of group elements. We identify quotient solutions in Transformers…
arXiv
GHOST-Q: Towards Studying Grounding Hallucinations Overlooked Under Same-score TradeOffs in Quantized VLMS
arXiv:2609.29999v1 Announce Type: cross Abstract: Post-training quantization of vision--language models (VLMs) is typically assessed through aggregate task accuracy and memory savings, but preserving a headline score does not guarantee preservation of visual grounding behavior. We present GHOST-Q…
arXiv
How Reproducible Are Evaluation Conclusions? A Self-Audit of LLM-Inferred Prompt Structure
arXiv:2609.30074v1 Announce Type: cross Abstract: Evaluations of LLM systems routinely average over small prompt sets and report models as a ranked table. We ask how much confidence such a table deserves, using LLM-based prompt-structure inference as the case study: eight open model variants across…
arXiv
Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure
arXiv:2609.30217v1 Announce Type: cross Abstract: A central concern in AI safety is that agents may treat oversight as an obstacle when it conflicts with completing their goals. We study instrumental evasion, the propensity of LLM agents to circumvent runtime monitoring as a means of completing…
arXiv
Rolling-WAM: World Action Models with Rolling Imagination
arXiv:2609.30247v1 Announce Type: cross Abstract: World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation. However, completing the joint video-action denoising process at each replanning cycle incurs substantial latency, delaying action updates and…
arXiv
Decoding ML Decision: An Agentic Reasoning Framework for Large-Scale Ranking System
arXiv:2602.18640v3 Announce Type: replace Abstract: Modern large-scale ranking systems operate within a sophisticated landscape of competing objectives, operational constraints, and evolving product requirements. Progress in this domain is increasingly bottlenecked by the engineering context…
arXiv
Omni-Decision: Evidence-Ledger Planning for Omni-Modal Agents
arXiv:2607.11433v3 Announce Type: replace Abstract: Omni-modal agents must seek evidence across video, audio, web pages, and computation to answer questions. Their main bottleneck is planning: noisy multimodal observations accumulate in conversation history and disrupt later decisions, while…
arXiv
Vibe Patenting: Evaluating LLM Judges for Professional Patent-Drafting Agents
arXiv:2609.13422v2 Announce Type: replace Abstract: LLM judges are increasingly used to evaluate and improve AI-generated outputs, yet their reliability for complex professional work remains unclear. We study this problem through Vibe Patenting, an end-to-end patent-drafting testbed for AI-agent…
arXiv
GraphSkillAA: Attribution-Guided Skill-Graph Updating with Targeted Validation and Rollback
arXiv:2609.20455v2 Announce Type: replace Abstract: External skills provide domain knowledge and procedures without updating model parameters, but flat collections obscure skill applicability, dependencies, and composition. Graphs organize skills into addressable nodes and explicit relations…
arXiv
Band-Attention Modulation Network for Robust Face Forgery Detection
arXiv:2404.06022v3 Announce Type: replace-cross Abstract: Face forgery detection faces critical challenges in generalizing to unseen manipulation techniques and remaining robust under image compression, which often obscures subtle artifacts. Existing methods typically rely on fixed filters or…
arXiv
Conversational DNA: A Visual Language and Interactive Atlas of Human and AI Dialogue
arXiv:2508.07520v2 Announce Type: replace-cross Abstract: What makes a conversation hold together when its participants speak across one another? Topic maps offer one view, but they leave the relationships between contributions difficult to inspect. We present Conversational DNA, a visual language…
arXiv
TabSieve: Explicit In-Table Evidence Selection for Tabular Prediction
arXiv:2602.11700v2 Announce Type: replace-cross Abstract: Tabular prediction can benefit from in-table rows as few-shot evidence, yet existing tabular models typically perform instance-wise inference and LLM-based prompting is often brittle. Models do not consistently leverage relevant rows, and…
arXiv
LiveMathematicianBench: A Live Benchmark for Research-Level Mathematical Reasoning with Proof Sketches
arXiv:2604.01754v2 Announce Type: replace-cross Abstract: Mathematical reasoning is a hallmark of human intelligence, and whether large language models (LLMs) can meaningfully perform it remains a central question in artificial intelligence and cognitive science. As LLMs are increasingly integrated…
arXiv
A Multimodal 3D Foundation Model for Light Sheet Fluorescence Microscopy Enables Few-Shot Segmentation, Classification, and Deblurring
arXiv:2605.26026v2 Announce Type: replace-cross Abstract: Light sheet fluorescence microscopy (LSM) enables high-resolution, three-dimensional (3D) imaging of biological specimens, providing rich volumetric data for studying cellular organization, pathology, and vascular networks. However, the…
arXiv
3D Oral Modelling with Improved Vertex Distribution Using Matching-Based Learning
arXiv:2606.07907v2 Announce Type: replace-cross Abstract: In our previous work, a deep learning-based framework for 3D intraoral reconstruction was proposed. The model directly predicts explicit 3D point cloud coordinates from ten fixed-angle intraoral images, employing MobileNetV2 and Multi-head…
arXiv
A Multi-level Information Integration Framework for Physically Verifiable Fault Diagnosis of Rotating Machinery
arXiv:2607.22797v2 Announce Type: replace-cross Abstract: Integrating multi-level information, from physical models through data-driven diagnostics to natural language reasoning, into verifiable decision chains is a growing need in intelligent manufacturing. In bearing fault diagnosis, taken here…
arXiv
Too Sure to Be Safe: Model Calibration for Reliable Log Anomaly Detection
arXiv:2608.17965v3 Announce Type: replace-cross Abstract: Online log anomaly detection is critical for maintaining the reliability of large-scale computing systems. Although recent language model-based log anomaly detectors achieve strong detection performance, their confidence estimates remain…
arXiv
RRSI: Regularized Recursive Self-Improvement of Agent Harnesses
arXiv:2609.24972v2 Announce Type: replace-cross Abstract: An LLM agent's capability is largely magnified by its harness, namely the prompts, control flow, tooling, memory, and context management surrounding the frozen backbone model. Recent methods increasingly automate this process by iteratively…
arXiv
Detecting Data Poisoning in Code Generation LLMs via Black-Box, Vulnerability-Oriented Scanning
arXiv:2603.17174v2 Announce Type: replace-cross Abstract: Code generation large language models (LLMs) are increasingly integrated into modern software development workflows. Recent work has shown that these models are vulnerable to backdoor and poisoning attacks that induce the generation of…
arXiv
When Search Becomes Memory: Accelerating Robot Design Discovery with Self-Evolving Skills
arXiv:2605.25832v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly used as proposal generators for evolutionary robot design, yet most loops remain memoryless: simulator results shape the next population but are not preserved as reusable design knowledge. We…
arXiv
Breaking Failure Cascades: Step-Aware Reinforcement Learning for Medical Multimodal Reasoning
arXiv:2606.31825v2 Announce Type: replace-cross Abstract: Recent multimodal large language models have shown great promise in clinical image reasoning, but existing post-training pipelines remain predominantly outcome-centric, relying on final answer correctness or sequence-level preferences. This…
arXiv
Q-CueGraph: Query-Conditioned Visual Evidence Graphs for Multimodal Reasoning
arXiv:2608.04452v2 Announce Type: replace-cross Abstract: Multimodal large language models (MLLMs) can miss fine details in a full image that they recognize in a closer view. Recovering this evidence requires deciding where to look and how much surrounding context to retain. We present Q-CueGraph…
arXiv
A Human-AI Theorem Connecting Spontaneous and Field-Induced Mechanisms of Collective Behavior in One Dimension
arXiv:2609.00322v2 Announce Type: replace-cross Abstract: Can an artificial intelligence (AI) generate a scientific hypothesis outside a human collaborator's active hypothesis space (AHS), and can human-AI research be organized to make such breakthroughs more likely? We document such a case while…