May 2026 — page 2 of 2
- SocialReasoning-Bench Exposes Major Gaps in AI Agents' Ability to Negotiate on Users' Behalf· Agentic AI
- Study reveals ChatGPT's mixed impact on how developers solve unfamiliar code problems· AI and Software Engineering
- New Knowledge Graph Benchmarks the Teaching Capabilities of AI Models in K-12 Education· AI
- Withdrawn arXiv Paper Proposes Framework for Standardizing AI Agent Benchmarks· AI
- New Method Catches Tool-Use Errors Before Execution, Boosting AI Agent Reliability· Agentic AI
- New ML Framework Tackles Bug-Fixing with Dual Reward System and Line-Level Credit Assignment· AI
- New Detection System Flags AI-Fabricated Citations in Research Papers· AI
- Study reveals how to strategically refine AI-generated research ideas under budget constraints· AI
- Formal Verification Brings AI Safety Guarantees to Framework Level, Not Alignment· AI
- New Framework Lets AI Systems Build Better Disease Models with Full Audit Trails· AI
- Gradient Signals Mislead LoRA Rank Tuning in Reinforcement Learning· AI Research
- New system helps small AI agents pick the right skills within budget constraints· Agentic AI
- New benchmark exposes how AI agents struggle with ambiguous real-world policies· AI
- Researchers Show Tool Calls in Language Models Can Be Read and Steered Before Generation· AI
- Researchers Release Bilingual Tatarstan Geography Dataset with AI Question-Answering System· AI
- Researchers Propose Direct Metrics for Measuring LLM Tool Use Efficiency Beyond Accuracy· AI
- LLM Coding Agents Struggle When Structural Rules Get Complex· AI
- Study reveals tiny fraction of Python packages drive ecosystem-wide impact· Open Source
- Google DeepMind's AlphaEvolve moves into production, optimizing TPUs and enterprise workflows· AI
- New Framework Detects LLM Security Flaws Hidden in GitHub Automation Workflows· AI Security
- UniPrefill accelerates AI model startup times across any architecture with block-wise sparsification· AI
- Security detection rules evolve messily, not steadily—study of Sigma and Splunk repositories reveals ongoing tradeoffs· Security
- Multi-Stage Prompting Dramatically Boosts AI Report Writing Reliability· AI
- Withdrawn Study Flags How AI Agent Repair Rankings Shift Based on Evaluator Settings· AI
- Uzbek Language Game Platform Pairs AI With Digital Wordnet Resource· Agentic AI
- Spectral diagnostics of attention heads can't detect information flow direction, new theory shows· AI
- New Method Fixes LLM Training to Prioritize Logical Accuracy Over Smooth-Sounding Output· AI
- New Benchmark Reveals How AI Coding Agents Hide Vulnerabilities Through Task Decomposition· AI Safety
- New Framework Improves RAG Systems by Verifying Whether Retrieved Evidence Actually Answers Questions· AI
- How AI Coding Agents Are Reshaping Software Development From Specification to Delivery· Agentic AI
- New technique pinpoints exactly where humans and AI take over in co-written text· AI
- New Defense System Protects AI Agents From Hidden Prompt Injection Attacks· AI Security
- Domain-Rich Dataset Boosts AI Vision Models for Eye Disease Detection· AI
- Audio-Language Models Struggle to Use Clinical Context for Dysarthric Speech Recognition· AI
- Researchers blend real and synthetic data to improve fake news detection· AI
- Researchers Propose Shadow Memory Defense to Protect AI Agents from Multi-Step Attacks· Agentic AI
- Smaller AI Model Matches Frontier LLMs on Agent Tasks, Cuts Token Use by 30%· Agentic AI
- New Benchmark Evaluates AI Models' Ability to Extract Data from Noisy Medical Reports· AI
- Simpler Models Beat Transformers for Parsing Low-Resource Languages· AI
- Phionyx: A new AI architecture for auditable, deterministic LLM behavior· Agentic AI
- Large Language Models Can Catch Hidden Refactoring Bugs That IDE Tools Miss· AI
- Researchers Release SignVerse-2M, a Massive Pose-Based Dataset of 55+ Sign Languages· AI
- New framework uses expert security patterns to automate software vulnerability fixes· AI Security
- Automated System Generates Cloud Incident Troubleshooting Guides in Seconds· AI
- Study maps how developers use AI chatbots for software design tasks· AI
- LLMs Can Temper Partisan News Bias—But Misjudge How Well They Do It· AI
- New technique cuts AI model fabricated package hallucinations by 88%· AI
- AI coding agents struggle to replicate materials science research, new benchmark finds· AI
- Researchers Propose Method to Identify Whether AI Augmented or Automated a Task· AI
- Agent Capsules: Adaptive Runtime Optimizes Multi-Agent LLM Pipeline Efficiency· Agentic AI
- System Learns Your Life Goals by Watching How You Use Your Computer· Human-Computer Interaction
- Study: Simpler Methods Beat LLMs for Linking Software Issues to Commits· AI
- Study reveals why local attention improves transformer performance beyond efficiency gains· AI
- Researchers tackle repetitive language generation in AI models with geometric state-space approach· Agentic AI