Hans Bethe

In the quiet harbor of the Local Bubble, a rarefied cavity of hot gas around the Sun, an explosion recently thundered. A supernova went off tens of parsecs away, and its echo still lingers in the neighborhood, reshaping the cosmic-ray landscape.

In the stream of charged particles flying toward us from the depths of the Galaxy, iron turned out to be the odd one out. The ATIC-2 instrument, lifted on a balloon over Antarctica, compared how the fraction of nuclei from sulfur to chromium relative to iron changes with energy.

A translator who learned a language from thousands of books can interpret a conversation without being distracted by noise in the hall. Similarly, a neural network trained on 127,137 years of model climate produces a winter forecast in a fraction of a second, faster and more accurately than bulky dynamical systems.

The orchestra plays a symphony, and you listen only to the first violin. That is roughly how river discharge forecasts have worked for decades: the entire El Niño signal is reduced to a single number calculated from a small rectangle in the Pacific Ocean. Yet the ocean is a vast ensemble, where heat flows from one region to another, and its full score contains far more information than a single note.

A fisherman in Queensland notices that in some years snapper are harder to catch, even though the nets and boat are the same. The reason hides not in the catch itself but in the water temperature several months earlier, when the fry were just hatching. A warm current arrived not for the whole year but only for a few weeks, and that short episode decided the fate of an entire fish generation.

A translator who gets distracted by every rustle in the room turns a one-hour meeting into two. That is how old statistical El Niño models worked: they picked out individual numbers from ocean maps but missed how a warm patch off Peru connects to a cool tongue near Australia. The new neural network behaves differently: it follows not every word but the meaning of the paragraph.

Imagine an interpreter in a huge hall who hears not only the person next to them but also a whisper in the far corner, and from those fragments reconstructs the entire conversation. That's how the CAS-Canglong neural network works: it picks up connections between pressure, wind, and heat across the whole planet and uses them to predict what the ocean will look like three months from now.

In the dense swarm of old stars known as the globular cluster M15, astronomers have spotted not just stars but layers of time. Some stars were born earlier, others later, and their chemical composition reveals their origin with the precision of an archaeologist reading the strata of an ancient settlement.

Measuring a particle’s energy at the Large Hadron Collider is like trying to catch a whisper in a roaring crowd. When protons collide, a flurry of particles is born, their signals blending into a cacophony. Old-school methods, like a noise-clogged ear, systematically underestimate the energy. Scientists built a neural network that works right inside the detector.

When a massive star exhausts its thermonuclear fuel, its core collapses, and the outer layers are ejected into space at speeds of thousands of kilometers per second. This is how a supernova remnant is born — an expanding cloud of shock-compressed gas spanning light-years.

Deep inside stars, light nuclei fuse, giving birth to helium and streams of neutrons. On Earth, we recreate this process by combining deuterium and tritium — special isotopes of hydrogen.

A fusion reactor resembles a giant boiler where hydrogen is 'cooked' into helium. The plasma inside is heated to 150 million degrees — ten times hotter than the Sun's core. But like any boiler, scale forms: harmful impurities that settle on the walls and quench the plasma.

Predicted by Wolfgang Pauli and named by Enrico Fermi, neutrinos are nearly elusive particles. Inside dying supernova giants and merging neutron stars, an immense number are born. Each particle has a flavor that can change.

The James Webb Space Telescope looked back to a time when the universe, just after the Big Bang, was only a billion years old, and found the 'Sleepy' galaxy. Unlike its neighbors, which shone with young stars, it looked tired and dim.

Astronomers read the history of galaxies by their portraits. Previously, they used either lengthy analysis of light by colors or quick filtered snapshots, losing the details of shape.

For two atomic nuclei to fuse, they must overcome their fierce electric repulsion — like trying to bring two angry bulls together. Normally, only monstrous energy does the trick, as in stellar cores, where Hans Bethe explained fusion back in the 1930s.

Our Sun emits streams of neutrinos — particles that barely interact with matter. Passing through the dense solar material, they change their type — scientists call this 'flavor.' It's like a voice that changes as it moves through a crowd: the thicker the crowd, the more it gets distorted.

Galaxies are like giant kitchens where, in stellar cores, light atoms are fused into heavy ones — much like flour and water becoming bread. When a star dies, it spills its accumulated riches into the interstellar gas, seasoning it with 'spices': oxygen, carbon, iron.

In the first weeks of an embryo's life, a dense crowd of cells crawls across its body. These cells will later form the facial bones, parts of the heart, and nerves. The crowd doesn't move randomly: a few leader cells take the front, and the rest follow them, like cyclists in a peloton. For a long time, it was believed that leaders are a special breed of cells, born with a drive to lead.

An ordinary neural network for quantum chemistry is like a student who learns from mistakes but keeps tripping over the same spot. It crawls toward the right answer, is sensitive to settings, and often loses stability. For the fluorine molecule, the new method sped up calculations by 17,900 times compared to the popular Adam method.

In the crowded center of the Milky Way, individual stars are hard to pick out, but that's exactly where Patchick 126 hides — a faint cluster that has long resisted confident classification. Astronomers pointed the infrared spectrograph IGRINS and the Hubble Space Telescope at it to study its chemistry and age in detail for the first time.

Some stars fizzle out like a dud firework: instead of a bright blast, just a cloud of smoke. Usually, a massive star ends with a supernova explosion, leaving a black hole. But sometimes there's no bang: the star collapses straight into a black hole, almost unnoticed.

Some superluminous supernovae show a bright oxygen imprint in their light and expand rapidly, while others do not. It seemed like different phenomena. But supernova PTF12dam broke the rules: as its brightness grew, its oxygen signature changed, shifting from one type to another.

When a massive star runs out of fuel, it collapses and explodes, scattering the material from its interior across space. The remnant of such an explosion, Cassiopeia A, has long been known to astronomers for its lopsided appearance: bright clumps of material are scattered unevenly.

Igniting a miniature sun is a long-held dream of humanity. Typically, scientists try to confine scorching plasma (a gas of charged particles) at temperatures of hundreds of millions of degrees. But plasma tends to escape magnetic traps, and confinement comes at enormous cost.

Inside an exploding supernova, an unimaginable number of neutrinos are born — ghost particles that pass through almost everything without a trace.

A light pulse in an optical fiber is like a starter's pistol shot: short, sharp, and unmistakable against the silence. It is exactly such pulses that the new artificial neuron, built on a silicon chip, produces.

Water in a stream bends around rocks, but the math of that motion hides in a black box: that's how ordinary neural networks work. They give an answer but don't say why. Engineers can't trust a prediction if they can't see its reasoning. Fluid dynamics now has a transparent model.

In a glass of water, ultrasound creates tiny bubbles. They grow and shrink, and sometimes they collapse sharply, like a popping soap bubble, but thousands of times faster. Near these bubbles float small particles — for example, nickel spheres the size of dust motes.

Porous zeolite minerals work like tiny sponges with holes of a strictly defined size. Through such holes, molecules of one kind pass while others get stuck: this is the basis for water purification, gas separation, and catalysis in petrochemistry.

The James Webb Space Telescope peered into the remnant of the supernova Cassiopeia A — a cloud of gas left behind after a massive star exploded. There, it spotted an intricate web of glowing filaments and a mysterious structure dubbed the 'Green Monster,' featuring nearly circular holes and rings.

Over the course of a day, the brain accumulates impressions the way a room accumulates stuff. At night, during deep sleep, it doesn't just rest: it tidies up: it sorts through the day's events and decides what to keep and what to toss. Scientists built a computer model of the brain that learns from a few examples during the day and 'replays' them at night, as if dreaming.

At the Large Hadron Collider, protons smash into each other like two trucks full of confetti. The debris, dozens of new particles, scatters in all directions. The detector records their tracks, and then the real work begins: piecing together a complete picture of the collision from those tracks. Previously, this was done by a set of rules devised by physicists.

Hidden in the dense swarm of stars at the center of the Milky Way lies the globular cluster Tonantzintla 2. It went unnoticed for decades, yet it holds the key to the earliest history of our galaxy.

When moving to a new apartment, a person doesn't relearn how to open doors or turn on lights. Experience from the old home transfers automatically: a kitchen is a place with a stove and a sink, a bedroom is a place with a bed. The brain doesn't store an exact map but a flexible schema that it overlays onto unfamiliar space. Scientists have built a computer model of this transfer. It has two levels.

Deep in space, there's a cloud of gas that looks like a necklace with bright beads. This is the Necklace Nebula — the remains of a star that once swelled up and swallowed its neighbor. At the center, a pair remains: a small, hot star and its bloated companion, rich in carbon.

In a dense gas cloud where a star heavier than the Sun is forming, there is a tiny source of radio waves. It works like a laser: methanol molecules amplify the radiation, and we see a bright spot. But this spot behaves strangely: its brightness jumps up and down within a matter of hours.

Every thought leaves a trace — a faint electrical blip on the scalp. A device with dozens of wires, resembling a swim cap, catches these traces and turns them into wavy lines. From those lines, you can guess what a person feels even if they stay silent. In a new study, three computer programs competed to see which could read such lines most accurately and determine whether the mood was bad, neutral, or good.

An ornithologist in the jungle hears a trill, but the bird is hidden in the foliage. It would be great to get its portrait right away — at least from a photo database. Usually, that requires thousands of pairs of “song recording + photo of the same bird,” and collecting them is expensive and time-consuming. The authors took a detour: they used two ready-made models.

Take a child's spinning top and set it on the table. It spins steadily, with all its parts moving in the same direction. Now toss a handful of beads onto the floor and blow on them: each bead will fly its own way. That's roughly how astronomers distinguish two types of large galaxies: 'spinning tops' (fast rotators) and 'beads' (slow ones with chaotic stellar motion).