Anthropic Claude Mythos 5 has designed a quantum circuit that uses 17 times fewer gates than an earlier design. The result highlights how artificial intelligence could help researchers reduce quantum computing costs by improving algorithms rather than relying solely on hardware advances.
Table of Contents
The experiment was conducted under the guidance of researcher Andi Gu as part of Anthropic’s STEM Fellows Program, in collaboration with researchers from Harvard University and QuEra Computing. The experiments that Anthropic Claude Mythos 5 underwent included eight circuit design experiments across fields such as chemistry, physics simulation, optimization, number theory, and numerical computing.
Also Read | Stablecoin Remittances Show 84% Cost Savings Over Traditional Transfers
Anthropic Claude Mythos 5 Improves Circuit Efficiency
Quantum computers perform computations through qubits through operations known as quantum gates. In contrast to classical bits that take a state of either zero or one, qubits can take states together through superposition. This feature can allow quantum computers to solve some scientific problems intractable for traditional computers.
However, current quantum hardware remains limited by qubit availability and environmental noise. Minor fluctuations result in errors, while more prolonged computations become more complicated. Thus, there is a need to find circuits with minimal operation and memory requirements. Hence, algorithm development is an essential step towards improving quantum computing.
Anthropic provided Claude plain language descriptions of eight computation tasks as well as an automatic system for the evaluation of their results. The model created circuits, evaluated and improved its solutions in cycles without any human intervention during them. According to the company, some of the results significantly surpassed the best-known solutions by an order of magnitude.
Claude Reduces Memory Requirements in Quantum Circuits
One major result involved modular inversion, a mathematical operation used in elliptic-curve arithmetic. This operation is relevant to quantum algorithms targeting the elliptic-curve discrete logarithm problem, which has implications for evaluating the security of elliptic-curve cryptography against sufficiently capable quantum computers.
A reversible algorithm for arithmetic was developed by Claude with the use of divsteps. The method makes it possible to simplify the numbers using consecutive conditional operations. Conversion of the procedure to a quantum circuit leads to the problem of memory, as the reversible calculation should save some amount of information that would allow it to restore previous steps instead of discarding unnecessary temporary results.
The model discovered a technique to restore branching choices from intermediate numerical values. The mathematical identities were discovered, the approach was tested with numerical simulation, and its correctness was proven using the Lean proof assistant. Thus, the circuit does not have to store the whole history of the choices made.
Anthropic Claude Mythos 5 Cuts Quantum Circuit Gates
The optimal elliptic curve circuit uses about 17 times fewer gates compared to the smallest previously existing design while using 12% more qubits. The finding shows an important engineering dilemma whereby decreasing the number of operations can lead to increasing the memory needs, both resources should be considered during design.
The Anthropic Company also presented other improvements. One optimization decreased the number of gates in a Reed–Solomon decoding circuit used in the DQI quantum optimization algorithm by about 12 times. It appears that AI-assisted design can be used to find possible reductions in different mathematical and scientific tasks; however, all designs should be analyzed individually.
These achievements do not imply that existing quantum computers can already conduct any necessary computation. There are serious technical problems related to hardware constraints, error correction, and operational stability. The elliptic-curve finding is important for cryptographers, but it does not prove that modern devices can break popular cryptography.
AI Could Accelerate Quantum Computing Progress
The research builds on years of progress in reducing the estimated resources needed for demanding quantum simulations. According to Anthropic, FeMoco, an iron-sulfur cluster associated with biological nitrogen fixation, was estimated in 2017 to require 50 trillion Toffoli gates for computing its ground-state energy.
Algorithmic innovations brought this number down to 340 million gates, which is more than 150,000 times lower than the initial number. Automated optimization of circuits can help bring this number even lower, although actual results will depend on hardware advancements and other factors in the future.
Also Read | Pump.fun Launches Perpetual Futures With Up to 40x Leverage
This article contains market analysis and price predictions. These are not guarantees. Crypto markets are volatile. Always DYOR. Not financial advice.
