Lifeboat Foundation

This paper proposes a new four-dimensional chaotic system that consists of two active magnetically controlled memristors. The dynamic characteristics of the system, including equilibrium points, Lyapunov exponent spectrum, bifurcation diagram, double-parameter Lyapunov exponent, and attractor basin, are analyzed. The results indicate that the Lyapunov exponents of the system undergo abrupt changes. The bifurcation diagrams reveal the occurrence of sudden cusp bifurcations, and the diverse manifestations of two-parameter Lyapunov exponents under different parameter combinations further underscore the system’s complexity and variability. This chaotic system also possesses an infinite number of equilibrium points and coexisting attractors, demonstrating multiple stable states.

From the dynamical point of view, most cognitive phenomena are hierarchical, transient and sequential. Such cognitive spatio-temporal processes can be represented by a set of sequential metastable dynamical states together with their associatedions: The state is quasi-stationary close to one metastable state before a rapidion to another state. Hence, we postulate that metastable states are the central players in cognitive information processing. Based on the analogy of quasiparticles as elementary units in physics, we introduce here the quantum of cognitive information dynamics, which we term “cognon”. A cognon, or dynamical unit of thought, is represented by a robust finite chain of metastable neural states. Cognons can be organized at multiple hierarchical levels and coordinate complex cognitive information representations.

Theories of the electrophysiology of language comprehension are mostly informed by event-related potential effects observed between condition averages. We here argue that a dissociation between competing effect-level explanations of event-related potentials can be achieved by turning to predictions and analyses at the single-trial level. Specifically, we examine the single-trial dynamics in event-related potential data that exhibited a biphasic N400–P600 effect pattern. A group of multi-stream models can explain biphasic effects by positing that each individual trial should induce either an N400 increase or a P600 increase, but not both. An alternative, single-stream account, Retrieval-Integration theory, explicitly predicts that N400 amplitude and P600 amplitude should be correlated at the single-trial level.

This month in 1,871, James Clerk demonstrated entropy with a thought experiment now known as’s Demon.

The American Physical Society is a nonprofit membership organization working to advance physics by fostering a vibrant, inclusive, and global community dedicated to science and society.

Disruptive innovations in technology, such as humanoid robots and electric vehicles, will lead to significant changes in labor, economy, and society, posing both opportunities and challenges for the future.

Questions to inspire discussion.

Continue reading “The Future of Technology: Impact on Labor, Economy, and Society” »

An immunotherapy drug given before surgery instead of chemotherapy meant that over ten times more patients with a certain genetic profile were cancer-free after surgery, according to clinical trial results presented by researchers at UCL and UCLH.

The findings, presented at the American Society of Clinical Oncology (ASCO) Annual Meeting 2024, are interim results from the NEOPRISM-CRC phase II clinical trial assessing whether the immunotherapy drug pembrolizumab can improve outcomes for patients with stage two or stage three MMR deficient/MSI-High bowel cancer. The trial was a collaboration among UCL, UCLH, the Christie NHS Foundation Trust in Manchester, St. James’s University Hospital in Leeds, University Hospital Southampton and the University of Glasgow.

Bowel cancer is the fourth most common cancer in the UK, with around 42,900 cases a year. Though still predominantly a cancer that affects older people, cases among the under 50s have been increasing in recent decades.

One Thousand and One Pairs: A “novel” challenge for long-context language models.

Marzena Karpinska, Katherine Thai, Kyle Lo, Tanya Goyal, Mohit Iyyer UMass NLP, Allen Institute, & Princeton June 2024 📃 Paper: https://arxiv.org/abs/2406.16264 👩‍💻 Code & Sample Data: https://github.com/notifications/beta/shelf

Contribute to marzenakrp/nocha development by creating an account on GitHub.

Update: China´s Moon Mission Returned Now Samples from the #Moon to #Earth. Why this is important, specially for the origin of life:

On June 1, China’s Chang’e-6 lander touched down in the South Pole-Atkin Basin — the largest, deepest, and oldest impact crater on the Moon. The probe almost immediately set to work drilling into the ground to collect about 2 kilograms of lunar material, which is already headed back to Earth, with a landing in Mongolia planned for June 25. It isn’t just planetary geologists who are excited at what the returning rocks and soil might reveal. If we’re lucky, the first samples from the lunar farside could also include some of the oldest fossils ever found.

The SPA basin, as it’s sometimes called, is the result of a gigantic impact that occurred between 4.2 and 4.3 billion years ago, at a time when the Moon and Earth were very close neighbors. The crater is roughly 2,500 kilometers (1,600 miles) in diameter and between 6.2 km and 8.2 km (3.9 to 5.1 mi) deep, encompassing several smaller craters like the Apollo basin, where Chang’e-6 landed, and Shackleton crater, parts of which lie in perpetual shadow.

Continue reading “How China’s Moon mission could reveal the origins of life on Earth” »

A new startup emerged out of stealth mode today to power the next generation of generative AI. Etched is a company that makes an application-specific integrated circuit (ASIC) to process “Transformers.” The transformer is an architecture for designing deep learning models developed by Google and is now the powerhouse behind models like OpenAI’s GPT-4o in ChatGPT, Antrophic Claude, Google Gemini, and Meta’s Llama family. Etched wanted to create an ASIC for processing only the transformer models, making a chip called Sohu. The claim is Sohu outperforms NVIDIA’s latest and greatest by an entire order of magnitude. Where a server configuration with eight NVIDIA H100 GPU clusters pushes Llama-3 70B models at 25,000 tokens per second, and the latest eight B200 “Blackwell” GPU cluster pushes 43,000 tokens/s, the eight Sohu clusters manage to output 500,000 tokens per second.