A new study published in Nature Communications found that it is impossible to draw a direct causal link between a specific artwork and the output of diffusion-based image generators like Midjourney an…
Liquid AI, an MIT spinout, released four updated LFM2.5 checkpoints on August 19, trained with quantization-aware distillation (QAD) to retain roughly 97% of their BF16 average benchmark performance w…
Researchers at MIT's Computer Science & Artificial Intelligence Laboratory (CSAIL) found that diffusion models trained on larger datasets exhibit 'attribution decay,' where individual training inputs …
A new MIT CSAIL study published in Nature Communications finds that individual training examples often have no measurable influence on any specific output of large generative image models, a phenomeno…
A peer-reviewed MIT CSAIL study published August 18 in Nature Communications found that diffusion-model outputs become harder to attribute to any single training image, person, or artist as training s…
MIT CSAIL and Tsinghua University researchers developed GeoPT, an AI pre-training method that learns physics from 1.3 million synthetic particle interactions, cutting labeled data needs by 60% and ach…
Rodney Brooks, the roboticist who co-founded iRobot and now serves as co-founder and CTO of Robust.AI, argued in an August 1st essay that technology forecasts conflate four distinct time scales: resea…
MIT CSAIL and Microsoft Azure Research have developed Murakkab, a resource-efficient serving system for AI agent workflows that cuts cloud costs by up to 4.3x without requiring changes to workflow log…
Turing Award winner Richard Sutton and co-founder Khurram Javed left John Carmack's Keen Technologies to launch Oak Lab, a Canadian AI startup building continuously learning agents based on Sutton's O…
MIT CSAIL and Toyota Research Institute have developed SceneSmith, a framework using three AI agents powered by GPT-5.2 to build detailed 3D virtual environments for robot training, generating scenes …
TinyAgents, a Rust-based recursive language model harness, was released on Hacker News. It implements the Recursive Language Model (RLM) execution model, enabling agents to call sub-agents, graphs to …
MIT CSAIL Alliances launched a new podcast called Building 32, hosted by Karen Given, featuring conversations with pioneers and researchers shaping AI, robotics, and computing. The podcast explores th…
Liquid AI released LFM 2.5-230M, a 230-million-parameter model that outperforms larger models on data extraction benchmarks, achieving 22.51 on CaseReportBench versus 13.83 for Qwen3.5-0.8B and 2.28 f…
Blueshoe, an AI-native law firm founded by Casey O'Grady and Kai Yee Wan, launched as a full-stack consumer and plaintiff-side firm, not a legal AI tool. The Y Combinator-backed startup aims to rebuil…
Researchers at MIT CSAIL and Harvard SEAS developed Collaborative Battleship, a language-based testbed, and collected the BattleshipQA dataset from over 40 human games to study how AI agents ask quest…
Researchers from MIT CSAIL, the National University of Singapore, A*STAR, and the Singapore-MIT Alliance for Research and Technology developed MeMo, a modular framework that boosts large language mode…
Researchers from the National University of Singapore, MIT CSAIL, A*STAR, and SMART have developed MEMO, a framework that trains a separate, small language model to internalize new knowledge from docu…