What Improves During Pretraining? Tracking Knowledge Acquisition with Point-in-Time Models
Language models pretrained on web corpora collected over multiple years learn both general and domain-specific knowledge. How…
News
Research & Science
Papers, breakthroughs and what the labs found this week.
591 stories · Every headline links to the original publisher. HapnGo curates and groups; we do not copy articles.
Rare, Relevant, and Still Wrong: How Memory Fails LLM Group Decisions
When an LLM group's discussion is compressed into a short memory for a final decision maker, what should the memory keep? Two…
I built this, feedback welcome [P]
It's a 50 hour course for beginners in AI/ML (obviously it's fully free). Starts out with linear algebra, derivatives, statistics…
Your Robot’s Safety Functions Already Work. What If the Input Lies?
This article is brought to you by VicOne. A mobile robot slows for someone in a hallway. A collaborative arm eases its movement…
Ψ-0.6: Adaptive Streams of Thought in a Unified Vision-Language World Model [R]
Chain-of-thought transformed AI reasoning, but it is limited to words. What's the physical-world analogue? Introducing Ψ-0.6, one…
Teachers other than EMA for distillation ? [R]
Hi guys, Self-distillation methods, see Self-Distillation Fine Tuning SDFT, usually use an EMA (Exponential Moving Average)…
Regarding training LLMs
Would training LLMs on either smaller batches of text first instead of the (current) large batches of text or on easier…
How AI Is Changing the Pace of Discovery for Millions Living With Dementia
AI summary C-BRAIN, a consortium of 18 institutions led by Dr. Randall Bateman that launched in November 2025, is working with NVIDIA to apply AI to Alzheimer's research, including unpublished and proprietary "dark data" analysed through a hybrid federated setup. It runs 64 on-site NVIDIA data center GPUs plus Microsoft Azure cloud nodes and plans to use BioNeMo and MONAI. Its open source tool OpenScientist finished analyses in minutes that scientists expected to take weeks or months, for under $10 per run, while a second tool, InsightEngine, checks claims against evidence. Worldwide, 57 million people live with dementia.
I Used AI to Fake a Scientific Paper. Here’s What Happened
Science has a spam problem. And the spam filter is getting clogged up. According to all reports, academic publishing is drowning…
OpenAI asks outside experts, including a Nobel laureate, to check proofs its AI produced in math, computer science, and physics
OpenAI says its internal AI models have cracked three major open questions in math, computer science, and physics. One of them is…
Industrial Intelligence: The next growth opportunity for manufacturing
Every manufacturer has its own industrial intelligence. Some is documented, and some resides in the knowledge of experienced…
Study finds AI models overuse patterns linked to reading difficulty
Researchers from New York University and the University of Texas at Austin analyzed language from 14 models across chat, coding…
Dense Representation Learning and Differentiable Shape Modeling for Spine Landmark Detection
Spine landmark detection from spinal X-ray is a key step for estimating spine deformation severity, for building biomechanical…
Self-Keyed In-Parameter Memorization for External Memory-Free Dialogue Recall in Frozen Language Models
A large language model maintains conversational content in an external storage that is outside the model and grows as the…
Algorithmic Monoculture in Machine Learning-based Site-Specific Fertilizer Recommendation
Machine learning recommendations are often evaluated one model at a time, despite deployment across interconnected decision…
Provenance Signatures for Low Rank Adaptations of Image Diffusion Models
Diffusion models have revolutionized the generation of realistic artificial images. Efficient techniques such as Low Rank…
Developing a Speech Language Model for Basque
Although the recent development of LLMs has paved the way for multi-modal interaction, minority languages face difficulties…
HiLo-RD: Cross-Budget Self-Distillation for Masked Diffusion Language Models
We propose HiLo-RD (High-to-Low Budget Representation Distillation) to improve masked diffusion language models (MDLMs) under a…
Natural Hard-Negative Mining from Positives for Instruction-Following Retrieval
Instruction-following information retrieval (IFIR) requires retrievers to rank documents by both a query and its…
Noise Steering: Exploiting Wake Steering and Acoustic Directivity for Noise-Constrained Wind Farm Control via Reinforcement Learning
Existing noise mitigation strategies for wind farms usually consist in reducing turbines power output, without accounting for the…
How to do method research when I feel I have no guidance on doing it? [D]
I am a 3rd year PhD student in a small lab working on ML in healthcare. My PI is new (we have no PhD graduates yet - lab is 4…
How Check Point accelerates vulnerability research with Daybreak
With OpenAI Daybreak, Check Point speeds up the research behind new security protections, enabling its analysts to turn…
ICE plan to buy robot dogs spurs protest from hundreds of Boston Dynamics employees
US robotics company stands by ICE deal as more than 300 employees warn of risk.
Free, daily
The HapnGo Briefing, in your inbox
One email a day with the AI stories that matter, why they matter, and one tool worth trying. No spam, one-click unsubscribe.