Your AI agent's skills are lying to you about why they work
Skills don't teach your agent much of anything, according to a new 8,135-trial study: only 4.5% of skill use is actual knowledge injection. The rest is mostly the agent using the skill file to stay on track. And the more skills you add, the worse it gets at finding the right one...
10 questions every AI leader should be able to answer in 2027
From kill switches to Chief AI Officer authority, these are the ten questions separating AI leaders with real answers from leaders about to get a very uncomfortable board question...
Large Language Models in Mental Health: A Systematic Review of Applications, Innovations, and Ethical Challenges
arXiv:2608.18080v1 Announce Type: new
Abstract: We present a review on the applications of large language models (LLMs) in health, e.g., social media analysis, clinical conversational agents, therapy support tools, prompt engineering, multimodal learning, and ethical considerations. We integrate fi...
What Can Artificial Intelligence Learn from Medicine? Generative Analogies and Reliable Machine Learning Systems
arXiv:2608.18186v1 Announce Type: new
Abstract: In the past few years, machine learning (ML) has been widely (and to an extent, successfully) implemented in medicine. However, uncertainties surrounding ML have made it difficult to establish the bases of its epistemic and methodological warrants. In...
Progressive Refinement: An Iterative Pseudo-Labeling Approach for Mandarin-English Code-Switching ASR
Code-switching (CS), alternating languages within the same utterance, poses significant challenges for automatic speech recognition (ASR) due to limited CS training data. This paper applies an iterative pseudo-labeling training approach to CS-ASR for the first time, demonstrating its effectiveness i...
Proactive Road Safety Intervention in Australia: Predicting Risky Driving Hotspots from Connected Vehicle Data
arXiv:2608.16913v1 Announce Type: new
Abstract: Road safety monitoring has historically been reactive, relying on crash-record analysis after fatalities and injuries have already occurred. Proactive identification of high-risk locations and dangerous driving behaviour before incidents occur is a cr...
FedPref: Federated Preference Learning for Structured Radiology Report Extraction
arXiv:2608.16971v1 Announce Type: new
Abstract: Radiology reports describe findings and locations in free text, but downstream search and analysis require these relations in a fixed schema. Learning this extraction requires labels that are unevenly distributed across institutions: smaller hospitals...
Geometry Is Not Robustness: A Trajectory-Level Study of PGD Evaluation
arXiv:2608.14594v1 Announce Type: new
Abstract: Projected Gradient Descent (PGD) is widely used to evaluate adversarial robustness, typically via final adversarial accuracy, which does not capture model behaviour throughout the attack. Recent work proposes trajectory-level diagnostics, such as loss...
FLOPs vs Real Work: The Importance of Replication in AI Efficiency Assessment
arXiv:2608.14550v1 Announce Type: new
Abstract: AI efficiency has recently taken the spotlight in both academy and industry due to massive model scales, high energy demands, and environmental costs. While reporting Floating Point Operations (FLOPs) is a traditional approach for assessing computatio...
Global AI Regulations for FAIR and Ethics in High-Risk Use Cases: A Comparative Review
arXiv:2608.14562v1 Announce Type: new
Abstract: AI governance is shifting from voluntary ethics to enforceable, risk-based regulation, yet cross-jurisdictional divergence creates compliance uncertainty for operators of high-stakes AI. We present a comparative matrix for the EU, US, and China that m...
GRPO Beyond English: A Large-Scale Study of GRPO in Non-English and Multilingual Settings
Reinforcement Learning with Verifiable Rewards (RLVR), often optimized with Group Relative Policy Optimization (GRPO), has become a central recipe for improving the reasoning capabilities of pretrained language models but current studies remain heavily English-centric. We conduct a large-scale empir...
MVICAD2: Multi-View Independent Component Analysis with Delays and Dilations
Machine learning techniques in multi-view settings face significant challenges, particularly when integrating heterogeneous data, aligning feature spaces, and managing view-specific biases. These issues are prominent in neuroscience, where data from multiple subjects exposed to the same stimuli are ...
A Specialized Semismooth Newton Method for Kernel-Based Optimal Transport
Kernel-based optimal transport (OT) estimators offer an alternative, functional estimation procedure to address OT problems from samples. Recent works suggest that these estimators are more statistically efficient than plug-in (linear programming-based) OT estimators when comparing probability measu...
Scientists turn DNA into a memory device that uses 100x less power
Researchers combined synthetic DNA with a semiconductor to create an ultra-low-power memory device capable of storing and processing information in the same place. The bio-hybrid technology could eventually help make AI systems and next-generation computers far more energy efficient.
Scientists tracked kids for 8 years — the screen time result was unexpected
An eight-year Finnish study found that children who spent more time on screens tended to show better cognitive processing as teenagers, challenging common assumptions about screen use. Researchers say the key may be balancing physical activity with screen activities that encourage learning, creativi...
arXiv:2608.13590v1 Announce Type: new
Abstract: XGBoost is a very popular and powerful method for prediction. It iteratively fits simple decision trees to the residuals of the previous step. An efficient and scalable implementation is available. The standard loss function for XGBoost is the quadrat...
Training-Free Knowledge Transfer Across Model Scales through Activation-Guided Pruning
arXiv:2608.13596v1 Announce Type: new
Abstract: Heterogeneous model fusion seeks to combine models that differ in tasks, initializations, architectures, or scales. We study an underexplored cross-scale setting: improving a small recipient language model with a stronger donor despite substantial arc...
How to build AI in the age of collaborative coding
I'm Steve, co-founder and CEO of Builder.io. and I want to talk about something that I think most teams are getting wrong right now, even the ones who've already bought into AI...
Which Site, and When: A Free-Satellite-Data Test of Himalayan Glacial Lake Bursts, Landslides, and Ice Floods
arXiv:2608.12422v1 Announce Type: new
Abstract: Two free satellite signals carry real information about glacial-lake outburst risk in the Nepal Himalaya: radar interferometry sees a moraine dam slowly sagging, and satellite weather marks the weeks when a primed lake is under stress. A companion fea...
MARCH: Scaling Recurrent Memory with Content-Routed State Anchors
arXiv:2608.12435v1 Announce Type: new
Abstract: Transformers owe much of their strong long-context retrieval capability to a token-level memory that grows with context length. This flexibility, however, incurs a quadratic computation complexity during training and a key--value cache that grows line...