MAROKO133 Update ai: Quantum engineers turn theory into working physics using exciton-driv

📌 MAROKO133 Eksklusif ai: Quantum engineers turn theory into working physics using

Research led by scientists at the Okinawa Institute of Science and Technology (OIST) and Stanford University has demonstrated a new approach to Floquet engineering using excitons rather than photons. Floquet engineering is a field of physics in which scientists attempt to design new materials by shining light on them.

This approach in modern science might sound like attempts at alchemy, and on the face of it, Floquet engineering is attempting alchemy, but it aims to modify the material’s quantum states.

A relatively new field, it rests on the theory that when a system is subjected to repeated external forces, its overall behavior is richer than the forces themselves.

To explain this, scientists often cite examples of a pendulum or a swing. In both these cases, the repeated external force, also known as a periodic drive, lifts the pendulum or swing to greater heights even though the object is only oscillating back and forth.

Using this principle, scientists aim to imbue exotic quantum properties into ordinary materials. 

How does this happen? 

In materials such as semiconductors, atoms are arranged in a tight lattice, while electrons are confined to specific energy levels or bands defined by the atoms’ structure. When light at a specific frequency is shone on the atom, the electromagnetic photons interact with the electrons, shifting their energy bands. 

By tuning the frequency and intensity of light, electrons can also be made to occupy hybrid bands, thereby altering the material’s properties. When the light source is switched off, the electrons return to their original energy bands, restoring the material’s properties. 

While this has been used to demonstrate Floquet effects, light couples weakly with matter, requiring very high frequencies to achieve hybridization.

“Such high energy levels tend to vaporize the material, and the effects are very short-lived. By contrast, excitonic Floquet engineering requires much lower intensities,” said Xing Zhu, PhD student at OIST, who was involved in the research. 

How can excitons help? 

Excitons are formed in semiconductors when electrons are excited from their valence band to a higher energy level, or the conduction band, by photons.

This leaves a positively charged hole in the valence band, and, along with the negatively charged electron, forms a quasiparticle called an electron-hole pair, which exists until the electron falls back into its valence shell. 

“Because the excitons are created from the electrons of the material itself, they couple much more strongly with the material than light,” explained Gianluca Stefanucci, professor at the University of Rome Tor Vergata, in a press release.

“And crucially, it takes significantly less light to create a population of excitons dense enough to serve as an effective periodic drive for hybridization.”

To investigate if excitons could be used to extract Floquet effects, the researchers at OIST first excited a semiconductor with a light drive. After measuring the energy levels of electrons, the researchers then dialled down the optical drive by an order of magnitude and measured the electron signal 200 femtoseconds later, to capture Floquet effects independent of the optical drive. 

“It took us tens of hours of data acquisition to observe Floquet replicas with light, but only around two to achieve excitonic Floquet – and with a much stronger effect,” said Vivek Pareek, postdoctoral fellow at California Institute of Technology, who was a graduate student at OIST when the work was done. 

The discovery is exciting not just because it moves away from light drives, but also because it opens up a wide range of excitation options, such as phonons (acoustic vibrations), plasmons (free-floating electrons), and magnons (magnetic fields), in the future. 

The research findings were published in the journal Nature Physics. 

🔗 Sumber: interestingengineering.com

📌 MAROKO133 Update ai: Nous Research's NousCoder-14B is an open-source coding

Nous Research, the open-source artificial intelligence startup backed by crypto venture firm Paradigm, released a new competitive programming model on Monday that it says matches or exceeds several larger proprietary systems — trained in just four days using 48 of Nvidia's latest B200 graphics processors.

The model, called NousCoder-14B, is another entry in a crowded field of AI coding assistants, but arrives at a particularly charged moment: Claude Code, the agentic programming tool from rival Anthropic, has dominated social media discussion since New Year's Day, with developers posting breathless testimonials about its capabilities. The simultaneous developments underscore how quickly AI-assisted software development is evolving — and how fiercely companies large and small are competing to capture what many believe will become a foundational technology for how software gets written.

type: embedded-entry-inline id: 74cSyrq6OUrp9SEQ5zOUSl

NousCoder-14B achieves a 67.87 percent accuracy rate on LiveCodeBench v6, a standardized evaluation that tests models on competitive programming problems published between August 2024 and May 2025. That figure represents a 7.08 percentage point improvement over the base model it was trained from, Alibaba's Qwen3-14B, according to Nous Research's technical report published alongside the release.

"I gave Claude Code a description of the problem, it generated what we built last year in an hour," wrote Jaana Dogan, a principal engineer at Google responsible for the Gemini API, in a viral post on X last week that captured the prevailing mood around AI coding tools. Dogan was describing a distributed agent orchestration system her team had spent a year developing — a system Claude Code approximated from a three-paragraph prompt.

The juxtaposition is instructive: while Anthropic's Claude Code has captured imaginations with demonstrations of end-to-end software development, Nous Research is betting that open-source alternatives trained on verifiable problems can close the gap — and that transparency in how these models are built matters as much as raw capability.

How Nous Research built an AI coding model that anyone can replicate

What distinguishes the NousCoder-14B release from many competitor announcements is its radical openness. Nous Research published not just the model weights but the complete reinforcement learning environment, benchmark suite, and training harness — built on the company's Atropos framework — enabling any researcher with sufficient compute to reproduce or extend the work.

"Open-sourcing the Atropos stack provides the necessary infrastructure for reproducible olympiad-level reasoning research," noted one observer on X, summarizing the significance for the academic and open-source communities.

The model was trained by Joe Li, a researcher in residence at Nous Research and a former competitive programmer himself. Li's technical report reveals an unexpectedly personal dimension: he compared the model's improvement trajectory to his own journey on Codeforces, the competitive programming platform where participants earn ratings based on contest performance.

Based on rough estimates mapping LiveCodeBench scores to Codeforces ratings, Li calculated that NousCoder-14B's improvemen t— from approximately the 1600-1750 rating range to 2100-2200 — mirrors a leap that took him nearly two years of sustained practice between ages 14 and 16. The model accomplished the equivalent in four days.

"Watching that final training run unfold was quite a surreal experience," Li wrote in the technical report.

But Li was quick to note an important caveat that speaks to broader questions about AI efficiency: he solved roughly 1,000 problems during those two years, while the model required 24,000. Humans, at least for now, remain dramatically more sample-efficient learners.

Inside the reinforcement learning system that trains on 24,000 competitive programming problems

NousCoder-14B's training process offers a window into the increasingly sophisticated techniques researchers use to improve AI reasoning capabilities through reinforcement learning.

The approach relies on what researchers call "verifiable rewards" — a system where the model generates code solutions, those solutions are executed against test cases, and the model receives a simple binary signal: correct or incorrect. This feedback loop, while conceptually straightforward, requires significant infrastructure to execute at scale.

Nous Research used Modal, a cloud computing platform, to run sandboxed code execution in parallel. Each of the 24,000 training problems contains hundreds of test cases on average, and the system must verify that generated code produces correct outputs within time and memory constraints — 15 seconds and 4 gigabytes, respectively.

The training employed a technique called DAPO (Dynamic Sampling Policy Optimization), which the researchers found performed slightly better than alternatives in their experiments. A key innovation involves "dynamic sampling" — discarding training examples where the model either solves all attempts or fails all attempts, since these provide no useful gradient signal for learning.

The researchers also adopted "iterative context extension," first training the model with a 32,000-token context window before expanding to 40,000 tokens. During evaluation, extending the context further to approximately 80,000 tokens produced the best results, with accuracy reaching 67.87 percent.

Perhaps most significantly, the training pipeline overlaps inference and verification — as soon as the model generates a solution, it begins work on the next problem while the previous solution is being checked. This pipelining, combined with asynchronous training where multiple model instances work in parallel, maximizes hardware utilization on expensive GPU clusters.

The looming data shortage that could slow AI coding model progress

Buried in Li's <a href="https://nousresearch.com/nouscoder-14b-a-co…

Konten dipersingkat otomatis.

🔗 Sumber: venturebeat.com