Explore documented scientific discoveries, original research and new scientific developments. Search the library by topic, category or date.
Scientists at McGill University have discovered a surprising new ability of immune cells that could change how we understand long-term immunity. Researchers found that T cells can actually sense how stiff their surrounding tissues are, using these physical signals to help determine whether they develop into long-lasting memory cells that protect against future infections.
ScienceDailyRead article →NASA’s Artemis I mission has revealed that cosmic radiation near the Moon is higher than in Earth orbit, but not by as much as scientists might expect. Tiny experimental detectors also uncovered surprising details about how radiation passes through the human body, offering valuable insights for future lunar exploration.
ScienceDailyRead article →Researchers have discovered a surprising hidden ability in Europe's familiar black-and-yellow fire salamanders: they glow green and cyan under ultraviolet light. The mysterious fluorescence comes mainly from glands in their skin and their toxic secretions, which can continue glowing for more than 24 hours after being released.
ScienceDailyRead article →A major new study reveals that autistic children’s brains process faces differently, and those differences may begin earlier than scientists previously thought. Researchers found that brain signals associated with recognizing faces become more specialized as neurotypical children grow older, but show much less change in autistic children.
ScienceDailyRead article →By probing its molecular structure, biologists are uncovering how a tangle of unstructured proteins turns into a tough, flexible thread. The post How Spiders Spin ‘Silk Dope’ Into Nature’s Toughest Material first appeared on Quanta Magazine
Quanta MagazineRead article →By tracking over 1 million cells, researchers created a detailed map of embryonic development in mice. It could provide development and disease insights.
MIT NewsRead article →NASA’s Artemis III SLS (Space Launch System) core stage is integrated alongside the twin solid rocket boosters on Oct. 28 inside the Vehicle Assembly Building’s High Bay 3 at NASA’s Kennedy Space Center in Florida. Technicians lifted the 212-foot rocket stage from the facility’s High Bay 2 using a one-of-a kind lifting beam designed […]
NASARead article →MIT researchers showed that the APOE4 gene variant suppresses molecular signaling needed for oligodendrocytes to myelinate neurons and also demonstrated a proof-of-concept fix.
MIT NewsRead article →Five years ago, Gavin Gardner was a Michigan State University engineering student researching next-generation drone technology as part of a NASA grant. Garder expected the research to be demanding – but he was surprised that navigating the financial side of the work presented an even greater challenge. Although Gardner was used to leading fellow students […]
NASARead article →MIT engineers designed a robotic enteroscope that can inch through the hard-to-reach small intestine, making biopsies and drug delivery easier.
MIT NewsRead article →So far, 2026 has seen a flood of successes for volunteers who do NASA science, exemplifying NASA’s commitment to gold-standard science—transparent, reproducible and intrinsically collaborative. The Backyard Worlds: Planet 9 project announced that their project volunteers have now discovered most of the known brown dwarfs, celestial objects that are similar to planets. The Zooniverse platform, […]
NASARead article →In this Webb image released on Oct. 22, 2026, we see a turbulent cauldron of star formation filled with dense clumps of gas. This region is named N159 and is located 161,000 light-years away in the Large Magellanic Cloud, one of the satellite galaxies orbiting the Milky Way galaxy. This region has a lower metallicity […]
NASARead article →Written by Lucy Lim, Planetary Scientist at NASA’s Goddard Space Flight Center Earth planning date: Friday, Oct. 16, 2026 Following the Basque Lakes drill campaign, Curiosity has returned to a spot just a few meters away from the “Chiu Chiu” block to conduct our next drill campaign. “Chiu Chiu” itself wasn’t a good candidate for […]
NASARead article →arXiv:2610.36309v1 Announce Type: new Abstract: Deep reinforcement learning (RL) algorithms have been observed to lose the ability to adaptively learn from new data, a phenomenon termed Loss of Plasticity (LoP). While many metrics have been shown to correlate with it, the causal mechanisms and the geometric conditions from which it can arise have not been clearly identified.
arXiv MLRead article →arXiv:2610.36280v1 Announce Type: new Abstract: Molecular 3D structures underpin simulations of thermodynamics, kinetics, crystallization, and nucleation, making their accurate prediction central to drug discovery and materials design. However, molecular geometry can change dramatically across environments. Existing models either ignore environmental effects or train separate models for each environment.
arXiv MLRead article →arXiv:2610.36255v1 Announce Type: new Abstract: The information bottleneck (IB) trades compression of an input $X$ against prediction of a target $Y$ through a multiplier $\beta=1/\lambda$, yet each operating point is usually served by training, or fully conditioning, a new encoder.
arXiv MLRead article →arXiv:2610.36247v1 Announce Type: new Abstract: Offline-to-online reinforcement learning uses offline pretraining to accelerate subsequent online learning. While many methods assume the offline dataset remains available throughout fine-tuning, there are many settings where this is not the case. Prior work attributes failures in this setting primarily to critic recalibration under distribution shift.
arXiv MLRead article →arXiv:2610.36245v1 Announce Type: new Abstract: Improving on a baseline requires exploring away from it, precisely where its performance may be lost. We propose DUET (Doubling Under Empirical Tests), a horizon-free phased mixing scheme for combinatorial semi-bandits with arbitrary finite action sets.
arXiv MLRead article →arXiv:2610.36204v1 Announce Type: new Abstract: Link prediction remains difficult in sparse graphs, where diffusion heuristics lack connectivity and geometric methods suffer spectral instability when coupling learned representations with diffusion operators.
arXiv MLRead article →arXiv:2610.36202v1 Announce Type: new Abstract: The Persona Selection Model (PSM) posits that post-training selects a persona the model learned during pre-training. Work on subliminal learning has shown that models can impart covert signals (like "love of owls") into semantically unrelated outputs like strings of numbers, and that students trained on these outputs inherit those traits.
arXiv MLRead article →arXiv:2610.36190v1 Announce Type: new Abstract: Graph Neural Networks (GNNs) are widely used in modern machine learning, yet most existing frameworks remain fundamentally static, limiting their applicability in Dynamic Data Driven Application Systems (DDDAS) where observations arrive sequentially under uncertainty.
arXiv MLRead article →arXiv:2610.36180v1 Announce Type: new Abstract: Dynamical coarse-graining of neural networks gives macroscopic observables and enables training to be predicted, and controlled in a state space far smaller than the parameter space. In this paper, we first define the learning operator following the training law of neural networks and establish a descent theorem for the learning operator, which serves to determine…
arXiv MLRead article →arXiv:2610.36139v1 Announce Type: new Abstract: On a chain of $N$ states, any local learning rule with bounded communication range needs $\Theta(N)$ rounds before the loss can influence the earliest parameters, and this latency is usually attributed to depth. We show that it is a property of the communication geometry rather than of depth.
arXiv MLRead article →arXiv:2610.36113v1 Announce Type: new Abstract: Grokking is a delayed transition from memorization to generalization in deep neural networks (NNs). We examine whether scaling and renormalization group (RG) theories can describe this dynamics, establishing a quantitative physical framework for neural generalization transitions.
arXiv MLRead article →