LLM as a Neural Architect: Controlled Generation of Image Captioning Models Under Strict API Contracts
arXiv:2512.14706v1 Announce Type: new
Abstract: Neural architecture search (NAS) traditionally requires significant human expertise or automated trial-and-error to design deep learning models. We present NN-Caption, an LLM-guided neural architecture search pipeline that generates runnable image-cap...
Autonomous Source Knowledge Selection in Multi-Domain Adaptation
arXiv:2512.14710v1 Announce Type: new
Abstract: Unsupervised multi-domain adaptation plays a key role in transfer learning by leveraging acquired rich source information from multiple source domains to solve target task from an unlabeled target domain. However, multiple source domains often contain...
A Bayesian latent class reinforcement learning framework to capture adaptive, feedback-driven travel behaviour
arXiv:2512.14713v1 Announce Type: new
Abstract: Many travel decisions involve a degree of experience formation, where individuals learn their preferences over time. At the same time, there is extensive scope for heterogeneity across individual travellers, both in their underlying preferences and in...
AgREE: Agentic Reasoning for Knowledge Graph Completion on Emerging Entities
Open-domain Knowledge Graph Completion (KGC) faces significant challenges in an ever-changing world, especially when considering the continual emergence of new entities in daily news. Existing approaches for KGC mainly rely on pretrained language models’ parametric knowledge, pre-constructed queries...
The Communication Complexity of Distributed Estimation
We study an extension of the standard two-party communication model in which Alice and Bob hold probability distributions ppp and qqq over domains XXX and YYY, respectively. Their goal is to estimate
Ex∼p,y∼q[f(x,y)]\mathbb{E}_{x \sim p, y \sim q}[f(x, y)]Ex∼p,y∼q[f(x,y)]
to within additive error ε...
A decision-theoretic characterization of perfect calibration is that an agent seeking to minimize a proper loss in expectation cannot improve their outcome by post-processing a perfectly calibrated predictor. Hu and Wu (FOCS’24) use this to define an approximate calibration measure called calibratio...
Unified Open-World Segmentation with Multi-Modal Prompts
Recent years have witnessed the rapid development of open-world image segmentation, including open-vocabulary segmentation and in-context segmentation. Nonetheless, existing methods are limited to a single modality prompt, which lacks the flexibility and accuracy needed for complex object-aware prom...
The Model Context Protocol (MCP) is genuinely useful. It gives people who develop AI tools a standardized way to call functions and access data from external systems. Instead of building custom integrations for each data source, you can expose databases, APIs, and internal tools through a common pro...
Most-Read: The Stanford HAI Stories that Defined AI in 2025
Readers wanted to know if their therapy chatbot could be trusted, whether their boss was automating the wrong job, and if their private conversations were training tomorrow's models.
Following the publication of his new book, Building Applications with AI Agents, I chatted with author Michael Albada about his experience writing the book and his thoughts on the field of AI agents. Michael’s a machine learning engineer with nine years of experience designing, building, and deployi...
Agent Lightning: Adding reinforcement learning to AI agents without code rewrites
By decoupling how agents work from how they’re trained, Agent Lightning turns each step an agent takes into data for reinforcement learning. This makes it easy for developers to improve agent performance with almost zero code changes.
The post Agent Lightning: Adding reinforcement learning to AI age...
Promptions helps make AI prompting more precise with dynamic UI controls
Promptions helps developers add dynamic, context-aware controls to chat interfaces so users can guide generative AI responses. It lets users shape outputs quickly without writing long instructions.
The post Promptions helps make AI prompting more precise with dynamic UI controls appeared first on Mi...
Scientists reveal a tiny brain chip that streams thoughts in real time
BISC is an ultra-thin neural implant that creates a high-bandwidth wireless link between the brain and computers. Its tiny single-chip design packs tens of thousands of electrodes and supports advanced AI models for decoding movement, perception, and intent. Initial clinical work shows it can be ins...
Using AI to analyze Google Street View images of damaged buildings across 16 states, Stanford researchers found that destroyed buildings in poor areas often remained empty lots for years, while those in wealthy areas were rebuilt bigger and better than before.
GigaTIME: Scaling tumor microenvironment modeling using virtual population generated by multimodal AI
Using AI-generated virtual populations, Microsoft researchers uncovered hidden cellular patterns that could reshape how we understand and treat cancer.
The post GigaTIME: Scaling tumor microenvironment modeling using virtual population generated by multimodal AI appeared first on Microsoft Research...
This tiny implant sends secret messages to the brain
Researchers have built a fully implantable device that sends light-based messages directly to the brain. Mice learned to interpret these artificial patterns as meaningful signals, even without touch, sight, or sound. The system uses up to 64 micro-LEDs to create complex neural patterns that resemble...
Ideas: Community building, machine learning, and the future of AI
As the Women in Machine Learning Workshop (WiML) marks its 20th annual gathering, cofounders, friends, and collaborators Jenn Wortman Vaughan and Hanna Wallach reflect on WiML’s evolution, navigating the field of ML, and their work in responsible AI.
The post Ideas: Community building, machine learn...
Reducing Privacy leaks in AI: Two approaches to contextual integrity
New research explores two ways to give AI agents stronger privacy safeguards grounded in contextual integrity. One adds lightweight, inference-time checks; the other builds contextual awareness directly into models through reasoning and RL.
The post Reducing Privacy leaks in AI: Two approaches to co...
#485 – David Kirtley: Nuclear Fusion, Plasma Physics, and the Future of Energy
David Kirtley is a nuclear fusion engineer and CEO of Helion Energy, a company working on building the world’s first commercial fusion power plant by 2028. Thank you for listening ❤ Check out our sponsors: https://lexfridman.com/sponsors/ep485-sc See below for timestamps, transcript, and to give fee...
The mall 0.2.0 update for R and Python introduces support for external LLM providers like OpenAI and Gemini. This version also features parallel processing for R users, the ability to run NLP on string vectors in Python, and a brand new cheatsheet.
The AI Agent Race Heats Up: Who’s Leading in 2025?
Autonomous AI agents – once a sci-fi concept – are rapidly becoming a mainstream reality. These agents don’t just chat; they plan, reason, and act across digital environments to achieve user goals independently. As we move into 2025, the race to build these agents is in full swing, with tech giants ...