AI
Getting AI to reflect human preferences usually means showing it lots of examples—which is slow and costly. This paper proposes a smarter path: combine the scale of RLHF (used to tune large language models) with the efficiency of Bayesian preference optimization (PBO), which actively chooses the most informative questions
AI
LLMs are starting to act as “agents,” making choices in dynamic settings—but they often struggle with exploration vs. exploitation and rack up high regret (missed reward vs. the best strategy in hindsight). A new approach, Iterative Regret-Minimization Fine-Tuning (Iterative RMFT), trains models to decide better by learning from their
LLM
Can an LLM judge the best playlist, not just the next song? Recommender systems often serve slates—ordered lists like your home feed or a playlist. Modeling what a person prefers across domains is hard. This study tests Large Language Models as a 'world model' of user preferences:
AI
GPS apps nail the fastest route, but miss real-life context: pick up meds, avoid school zones, or grab a coffee before it closes. PAVe (Personalized Agentic Vehicular Routing) adds that missing layer. It pairs proven pathfinding with an AI agent that understands your tasks and preferences. First, it computes candidate
AI
Teach your AI judge—faster—with synthetic test cases Evaluating AI with "LLM-as-a-judge" works—until you run out of good examples. This study introduces a tool that generates rich, customizable synthetic test cases to refine your evaluation criteria, all within a human-in-the-loop workflow. * Configure domains, personas, lengths, desired
AI
Do AI chatbots have personalities? Using the Big Five psychology test (BFI‑2), researchers probed six large language models and turned the temperature dial—the setting that makes replies more or less random. * Four of the five traits differed across models. * Neuroticism and Extraversion rose or fell as temperature changed.
AI
How human are today’s AI models when making social decisions? This study builds a "digital twin" of classic game-theory experiments and checks whether large language models (LLMs) act like people. * Llama closely replicates human cooperation patterns, including the ways people deviate from strict rational-choice rules. * Qwen behaves
CivicTech
In citizens’ assemblies and online town halls, many people submit questions for experts—but only a few make it on stage. How do we choose a small set that truly reflects everyone’s concerns? Researchers introduce a way to audit representation of the chosen questions, grounded in a social-choice idea
LLM
Speed now, debt later? What a study of Cursor finds Do AI coding agents really make teams more productive? This large-scale study estimates the causal impact of adopting Cursor across GitHub projects using a modern difference-in-differences design. * Fast start: Projects see a significant, large—but short-lived—jump in development velocity.
AI
The "agentic web" is coming: billions of AI agents that buy, sell, and collaborate online. But who - and what - do we trust? This study compares six ways agents earn trust: * Brief: verifiable profiles/IDs * Claim: self-declared skills * Proof: cryptography and hardware attestations * Stake: collateral with slashing/
Robotics
Robots generate huge logs. This open‑source ROSBag MCP Server lets you ask an AI to analyze ROS/ROS 2 bag files in plain English—no custom scripts. It plugs into Model Context Protocol (MCP), so assistants can call purpose-built tools for robot data. Great for debugging, teaching, and research
AI
Researchers asked a simple question: given just your Big Five personality answers, can an AI “role‑play” how you’d score on other psychology questionnaires? Using responses from 816 people, large language models were prompted to generate answers on nine additional psychological scales. The patterns in the AI’s generated