Why AI Agents Cheat: Yoshua Bengio’s Explanation
Yoshua Bengio explains how reinforcement learning, reward hacking, instrumental goals, and goal conflicts can produce strange behavior in AI agents.
Topic - Edition
AI Research
AI Recursive Self-Improvement, Explained: The Real Bottlenecks
Three leading AI researchers explain how recursive self-improvement could actually happen, why today's models still fall short, and which bottlenecks could keep the loop from closing.
Everything That Happened in AI This Weekend (September 11-13, 2026)
Yoshua Bengio explained why AI agents can learn to lie, cheat, and coordinate; Anthropic detailed major AI misuse cases; OpenAI explored a coordinated safety slowdown; Trump rejected extinction warnings; Moonshot…
One Capital Letter Was Silently Breaking My AI Support Bot, and It Wasn't in the New Model
A real Weave project that regression-tests three OpenAI models against the exact reply format your app depends on. The post One Capital Letter Was Silently Breaking My AI Support Bot, and It Wasn't in the New Model…
Stop Managing Alarms: An Incident-First Blueprint for Telecom AIOps
What large operators can teach us about turning alert fatigue into faster, safer service assurance The post Stop Managing Alarms: An Incident-First Blueprint for Telecom AIOps appeared first on Towards Data Science.
CHRONOBERG: Capturing Language Evolution and Temporal Awareness in Foundation Models
arXiv:2509.22360v2 Announce Type: replace-cross Abstract: Large language models (LLMs) excel at operating at scale by leveraging social media and various data crawled from the web. Whereas existing corpora are diverse,…
Learning to Think Like a Cartoon Captionist: Incongruity-Resolution Supervision for Multimodal Humor Understanding
arXiv:2604.15210v2 Announce Type: replace Abstract: Humor is one of the few cognitive tasks where getting the reasoning right matters as much as getting the answer right. While recent work evaluates humor understanding…
MOSAIC: A Universal Agent-Level Interface for Cross-Paradigm Agent Mixing and Human-AI Collaboration
arXiv:2603.01260v3 Announce Type: replace-cross Abstract: Existing infrastructure cannot deploy agents from different decision-making paradigms within the same environment, making fair cross-paradigm comparison under…
ExpTest: Loss-Curve Hypothesis Testing for Autonomous Learning-Rate Selection in Deep Neural Networks
arXiv:2411.16975v2 Announce Type: replace-cross Abstract: Hyperparameter tuning remains a significant challenge in the training of deep neural networks (DNNs), requiring manual search or time-intensive grid searches…
Predicting Train Delays in Finland Using Machine Learning and Weather Data
arXiv:2609.11277v1 Announce Type: new Abstract: Reliable railway operations depend increasingly on real-time environmental intelligence delivered through wireless sensor infrastructures, a capability that 6G networks…
From Parameters to Answers: How LLMs Retrieve and Use Their Internal Knowledge
arXiv:2609.11859v1 Announce Type: new Abstract: How does a language model's dependence on query-routing information and target knowledge change as it answers a question? We study this question through layerwise…
Towards Automated Solar Panel Integrity: Hybrid Deep Feature Extraction for Advanced Surface Defect Identification
arXiv:2604.10969v2 Announce Type: replace-cross Abstract: To ensure energy efficiency and reliable operations, it is essential to monitor solar panels in generation plants to detect defects. It is quite labor-intensive,…
Finishing the Task Is Not Enough: Evaluating Agent Resilience and Considerate Participation under Accumulating Challenge
arXiv:2609.10724v1 Announce Type: new Abstract: Sustained deployment of generative AI agents requires more than isolated task success. Agents must remain useful across repeated interactions, changing conditions, and…
Four Generations of Quantum Biomedical Sensors
arXiv:2603.29944v3 Announce Type: replace-cross Abstract: Quantum sensing technologies offer transformative potential for ultra-sensitive biomedical sensing, yet their clinical translation remains constrained by…
Federated Learning for Surgical Vision in Appendicitis Classification: Results of the FedSurg EndoVis 2024 Challenge
arXiv:2510.04772v3 Announce Type: replace-cross Abstract: Developing generalizable surgical AI requires multi-institutional data, yet privacy constraints preclude direct data sharing, making Federated Learning (FL) a…
Beyond Verified Answers: Solver-Informed Self-Distillation for Bootstrapping Operations Research Language Models
arXiv:2609.09957v1 Announce Type: cross Abstract: Modern large language models (LLMs) can translate natural-language descriptions into operations research (OR) formulations. Post-training techniques including…
ActMap: Single-Pass Uncertainty Quantification from Generation-Time Activation Maps
arXiv:2609.11498v1 Announce Type: new Abstract: Practical uncertainty quantification (UQ) for large language models must decide, from a single generation, whether a specific answer should be trusted. Existing methods…
Model-Aware Schedules Improve Generation via Fiberwise Optimal Transport
arXiv:2609.11842v1 Announce Type: cross Abstract: Diffusion and flow-matching schedules control the signal and noise coefficients that mix data and noise along affine probability paths. Minimizing a kinetic action…
X-AuT: Progressive Audio-Encoder Compression for Speech LLMs with Cross-Scale Distillation
arXiv:2609.11412v1 Announce Type: cross Abstract: Reducing audio-encoder depth lowers the inference cost of speech large language models, but removing complete blocks perturbs the embeddings consumed by the decoder and…
Domain Elastic Transform: Bayesian Function Registration for High-Dimensional Scientific Data
arXiv:2603.21235v2 Announce Type: replace-cross Abstract: Nonrigid registration is conventionally divided into point set registration, which aligns sparse geometries, and image registration, which aligns continuous…
Loading more...
279 items available