TL;DR
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
Researchers from Anthropic and Harvard report using Claude to design more resource-efficient circuits for eight quantum-computing tasks. Their estimates include a decoder with about 440,000 Toffoli gates, compared with 5.1 million in the study’s baseline; the designs were checked computationally, not demonstrated on quantum hardware.
Researchers from Anthropic and Harvard say they used Claude to design more resource-efficient quantum circuits across eight computational tasks, including a decoder estimated to need about 12 times fewer costly Toffoli gates than the study’s comparison design. The results, posted as an arXiv preprint, are based on computational checks and resource estimates—not demonstrations on a working quantum computer.
In the decoder example, the proposed circuit uses about 440,000 Toffoli gates, down from 5.1 million in the baseline. Its estimated peak qubit requirement also falls, from 1,938 to 1,837. Toffoli gates are a costly operation in many proposed error-corrected quantum-computing systems, so their number is one measure of the work a circuit may require.
The team reports that seven of the eight designs reduced both gate counts and qubit requirements relative to the study’s chosen baselines. The eighth was a new arithmetic design that gives developers different trade-offs between computation time and memory. Those findings concern component circuits, not complete algorithms running on quantum machines.
According to Anthropic’s accompanying account, the researchers equipped Claude with programming tools and an evaluator that checked proposed circuits and scored their resource estimates. The model revised its proposals using that feedback. Anthropic identifies the model as Claude Mythos 5 and says its elliptic-curve arithmetic work took 53 hours. These details are company-reported; the study’s practical performance claims remain subject to the limits of estimates and the preprint’s methods.
Lower Costs for Quantum Algorithms
Quantum algorithms are built from smaller routines for arithmetic, data access and numerical calculations. If those components require fewer operations or qubits, they could reduce the hardware and error-correction demands of a larger computation. That matters because useful quantum computing depends not only on building larger, more reliable machines, but also on making the algorithms they run more efficient.
The decoder result illustrates both the potential and the limits. A reduction from 5.1 million to about 440,000 Toffoli gates is substantial within the comparison used by the researchers. It does not, by itself, establish that a complete quantum calculation will run faster or become feasible: overhead elsewhere in the algorithm, fault-tolerance requirements and hardware constraints can change the total resource bill. The work is also an example of AI being used as a search and design aid, with its outputs checked by computational tools rather than accepted without verification.
quantum computing circuit simulator
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
How the Circuit Search Worked
A quantum circuit is a sequence of operations on qubits. Designing one involves constraints that do not arise in the same way in ordinary software. A quantum operation generally has to preserve enough information to be reversed, so temporary calculations cannot simply be overwritten. Designers may need to keep intermediate results in extra qubits or later erase them through uncomputation, which can add more operations.
The preprint describes a search process in which the model could make and test circuit designs against resource estimates. The researchers report strategies such as recovering information from the current calculation state instead of storing a record of every earlier step, and sharing arithmetic across computational paths. In the decoder case, the design uses the Berlekamp–Massey method in place of an approach based on the extended Euclidean algorithm. The paper also describes limiting resources for rare inputs, accepting a small probability of failure in the relevant task.
The decoder is part of decoded quantum interferometry, an algorithmic approach that aims to make better solutions more likely when a quantum state is measured. In this study, the routine addresses a curve-fitting problem. Its error-detection role is within that algorithm; it should not be confused with correcting physical faults in quantum-computing hardware.
“The elliptic-curve arithmetic work took 53 hours.”
— Anthropic’s accompanying blog post
As an affiliate, we earn on qualifying purchases.
Hardware Tests Still Needed
The circuits have not been shown running on a working quantum computer, and the reported resource reductions are estimates based on computational checks. It remains unclear how much of the savings would survive when each component is integrated into a complete algorithm and mapped to a particular machine’s architecture and error-correction scheme.
The decoder design also makes a trade-off involving rare inputs. The paper says a storage requirement that could reach 665 qubits in the worst case was replaced by an allocation of 19 qubits, with the design accepting failure for inputs that exceed that capacity. The supplied account says a second stored entry would be needed with a probability of about one in 10,000, but does not provide enough detail here to determine the full effect of these probabilities on an end-to-end computation. The model’s reported contribution and the researchers’ verification process also do not establish that the same results will generalize to other tasks.
As an affiliate, we earn on qualifying purchases.
From Estimates to Full Algorithms
The next test is whether the proposed circuits can be incorporated into complete quantum algorithms while retaining their estimated savings. That will require evaluating the designs against realistic hardware and fault-tolerance assumptions, and establishing how their resource costs compare with alternatives under the same conditions.
The work is currently available as a preprint, so its methods and results have not yet been established through peer review. Further scrutiny of the circuit constructions, failure probabilities and baseline choices will help show how robust the comparisons are. The supplied report does not give a timetable for hardware demonstrations or identify a scheduled next research milestone.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did the researchers report?
Anthropic and Harvard researchers report that Claude helped design circuits for eight quantum-computing tasks. They say seven reduced both estimated costly-operation counts and qubit requirements against the study’s baselines, while the eighth offered a different arithmetic trade-off.
Was the circuit demonstrated on quantum hardware?
No hardware demonstration is reported. The findings come from computational checks and resource estimates, so performance on a working machine remains unconfirmed.
How much smaller was the decoder circuit?
The study estimates about 440,000 Toffoli gates for the redesigned circuit, compared with 5.1 million in its baseline. Estimated peak qubits fell from 1,938 to 1,837.
Does the result mean quantum computers can now solve more problems?
Not on its own. Leaner components may lower resource demands, but the reported work does not establish that a complete algorithm can run on current hardware or that it will solve a problem otherwise out of reach.
Has the research been peer-reviewed?
The source describes the work as an arXiv preprint. The supplied material does not report peer review or a hardware test.
Source: rss
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
