Zoe Holmes
@qzoeholmes.bsky.social
Quantum physicist. Assistant Prof at EPFL. Climber.
Please accept my sincere apologies for this notation.
(Though I also challenge you to come up with anything less headache inducing in this context).
I don’t really know what you mean by ’best randomness’ but your intuition is probably in the right direction.
(Though I also challenge you to come up with anything less headache inducing in this context).
I don’t really know what you mean by ’best randomness’ but your intuition is probably in the right direction.
November 22, 2025 at 5:21 PM
Please accept my sincere apologies for this notation.
(Though I also challenge you to come up with anything less headache inducing in this context).
I don’t really know what you mean by ’best randomness’ but your intuition is probably in the right direction.
(Though I also challenge you to come up with anything less headache inducing in this context).
I don’t really know what you mean by ’best randomness’ but your intuition is probably in the right direction.
So we shook hands, agreed and lived happily ever after
Oh and we also derived a shit tonne of expressions for moments of channel ensembles - so if you ever need these you know where to look!
Thanks for the (heated) project @mduschenes.bsky.social @dgarciamartin.bsky.social @mvscerezo.bsky.social
Oh and we also derived a shit tonne of expressions for moments of channel ensembles - so if you ever need these you know where to look!
Thanks for the (heated) project @mduschenes.bsky.social @dgarciamartin.bsky.social @mvscerezo.bsky.social
November 21, 2025 at 10:13 PM
So we shook hands, agreed and lived happily ever after
Oh and we also derived a shit tonne of expressions for moments of channel ensembles - so if you ever need these you know where to look!
Thanks for the (heated) project @mduschenes.bsky.social @dgarciamartin.bsky.social @mvscerezo.bsky.social
Oh and we also derived a shit tonne of expressions for moments of channel ensembles - so if you ever need these you know where to look!
Thanks for the (heated) project @mduschenes.bsky.social @dgarciamartin.bsky.social @mvscerezo.bsky.social
And so I conceded that @mvscerezo.bsky.social had a good point
Perhaps the ensemble containing just the completely depolarizing channel is the most natural generalization of Haar
(But I wanted to drop the word `expressivity' as that has connotations with being useful/varied & this ensemble is not)
Perhaps the ensemble containing just the completely depolarizing channel is the most natural generalization of Haar
(But I wanted to drop the word `expressivity' as that has connotations with being useful/varied & this ensemble is not)
November 21, 2025 at 10:13 PM
And so I conceded that @mvscerezo.bsky.social had a good point
Perhaps the ensemble containing just the completely depolarizing channel is the most natural generalization of Haar
(But I wanted to drop the word `expressivity' as that has connotations with being useful/varied & this ensemble is not)
Perhaps the ensemble containing just the completely depolarizing channel is the most natural generalization of Haar
(But I wanted to drop the word `expressivity' as that has connotations with being useful/varied & this ensemble is not)
I kept arguing that cHaar was the natural generalization
But then we found:
1) Concatenating cHaar made it tend towards the depolarising channel
2) As did increasing the size of the environment of cHaar
It was like cHaar with a finite environment wasn't fully uniform but could be made more so...
But then we found:
1) Concatenating cHaar made it tend towards the depolarising channel
2) As did increasing the size of the environment of cHaar
It was like cHaar with a finite environment wasn't fully uniform but could be made more so...
November 21, 2025 at 10:13 PM
I kept arguing that cHaar was the natural generalization
But then we found:
1) Concatenating cHaar made it tend towards the depolarising channel
2) As did increasing the size of the environment of cHaar
It was like cHaar with a finite environment wasn't fully uniform but could be made more so...
But then we found:
1) Concatenating cHaar made it tend towards the depolarising channel
2) As did increasing the size of the environment of cHaar
It was like cHaar with a finite environment wasn't fully uniform but could be made more so...
The ensemble that satisfied these properties contains only one channel - the complete depolarizing channel.
Based on this @mvscerezo.bsky.social argued the ensemble was maximally expressive.
I thought this crazy.
How could this ensemble - where you throw away your state - be maximally expressive?
Based on this @mvscerezo.bsky.social argued the ensemble was maximally expressive.
I thought this crazy.
How could this ensemble - where you throw away your state - be maximally expressive?
November 21, 2025 at 10:13 PM
The ensemble that satisfied these properties contains only one channel - the complete depolarizing channel.
Based on this @mvscerezo.bsky.social argued the ensemble was maximally expressive.
I thought this crazy.
How could this ensemble - where you throw away your state - be maximally expressive?
Based on this @mvscerezo.bsky.social argued the ensemble was maximally expressive.
I thought this crazy.
How could this ensemble - where you throw away your state - be maximally expressive?
We then starting studying these ensembles and found that they didn't inherit a number of nice properties that the Haar measure has.
1) They were not left and right invariant.
2) The 2-norm of their moment operator is not minimal.
And in fact another ensemble satisfies these properties...
1) They were not left and right invariant.
2) The 2-norm of their moment operator is not minimal.
And in fact another ensemble satisfies these properties...
November 21, 2025 at 10:13 PM
We then starting studying these ensembles and found that they didn't inherit a number of nice properties that the Haar measure has.
1) They were not left and right invariant.
2) The 2-norm of their moment operator is not minimal.
And in fact another ensemble satisfies these properties...
1) They were not left and right invariant.
2) The 2-norm of their moment operator is not minimal.
And in fact another ensemble satisfies these properties...
The seemingly natural option is to use the Stinespring dilation and consider the uniform distribution over the channel purification.
I.e. to say the `maximally uniform' distribution of channels is the Haar distribution with the environment traced out.
We called this ensemble cHaar: Channel Haar
I.e. to say the `maximally uniform' distribution of channels is the Haar distribution with the environment traced out.
We called this ensemble cHaar: Channel Haar
November 21, 2025 at 10:13 PM
The seemingly natural option is to use the Stinespring dilation and consider the uniform distribution over the channel purification.
I.e. to say the `maximally uniform' distribution of channels is the Haar distribution with the environment traced out.
We called this ensemble cHaar: Channel Haar
I.e. to say the `maximally uniform' distribution of channels is the Haar distribution with the environment traced out.
We called this ensemble cHaar: Channel Haar
It begins in 2022 at the LANL summer school
We aimed to generalize `expressivity' from parameterized quantum circuits to quantum channels
Since unitary expressivity is measured via the distance of the ensemble from the Haar distribution, we wanted a channel generalization of the Haar measure.
We aimed to generalize `expressivity' from parameterized quantum circuits to quantum channels
Since unitary expressivity is measured via the distance of the ensemble from the Haar distribution, we wanted a channel generalization of the Haar measure.
November 21, 2025 at 10:13 PM
It begins in 2022 at the LANL summer school
We aimed to generalize `expressivity' from parameterized quantum circuits to quantum channels
Since unitary expressivity is measured via the distance of the ensemble from the Haar distribution, we wanted a channel generalization of the Haar measure.
We aimed to generalize `expressivity' from parameterized quantum circuits to quantum channels
Since unitary expressivity is measured via the distance of the ensemble from the Haar distribution, we wanted a channel generalization of the Haar measure.
Sit down & buckle up.
Behind this rather innocent title is the paper that caused the most fraught arguments of any of my career.
Are these arguments important? Probably not.
Am I going to tell you them anyway? Absolutely.
Here is the story of a rollercoaster of a paper.
🎢👇
Behind this rather innocent title is the paper that caused the most fraught arguments of any of my career.
Are these arguments important? Probably not.
Am I going to tell you them anyway? Absolutely.
Here is the story of a rollercoaster of a paper.
🎢👇
November 21, 2025 at 10:13 PM
Sit down & buckle up.
Behind this rather innocent title is the paper that caused the most fraught arguments of any of my career.
Are these arguments important? Probably not.
Am I going to tell you them anyway? Absolutely.
Here is the story of a rollercoaster of a paper.
🎢👇
Behind this rather innocent title is the paper that caused the most fraught arguments of any of my career.
Are these arguments important? Probably not.
Am I going to tell you them anyway? Absolutely.
Here is the story of a rollercoaster of a paper.
🎢👇
Reposted by Zoe Holmes
(I/III) We're excited to announce a new tenure track opening! The position is called 'quantum informatics' and is affiliated with our QUICK group within the CS+AI division at @jku.at 🇦🇹. Application deadline is November 30th, 2025: www.jku.at/en/the-jku/w...
October 21, 2025 at 1:47 PM
(I/III) We're excited to announce a new tenure track opening! The position is called 'quantum informatics' and is affiliated with our QUICK group within the CS+AI division at @jku.at 🇦🇹. Application deadline is November 30th, 2025: www.jku.at/en/the-jku/w...
Applications for the LANL quantum computing summer school are now open!
I recently got asked about the most important turning points of my career and without hesistation said the LANL summer school.
The first 1/3 of this video is me rambling about why it was so special: youtu.be/XjkHmtr_IT0?...
I recently got asked about the most important turning points of my career and without hesistation said the LANL summer school.
The first 1/3 of this video is me rambling about why it was so special: youtu.be/XjkHmtr_IT0?...
November 18, 2025 at 4:49 PM
Applications for the LANL quantum computing summer school are now open!
I recently got asked about the most important turning points of my career and without hesistation said the LANL summer school.
The first 1/3 of this video is me rambling about why it was so special: youtu.be/XjkHmtr_IT0?...
I recently got asked about the most important turning points of my career and without hesistation said the LANL summer school.
The first 1/3 of this video is me rambling about why it was so special: youtu.be/XjkHmtr_IT0?...
Reposted by Zoe Holmes
Our group is looking to hire postdocs to work at the intersection of quantum computing, quantum algorithms, quantum machine learning, simulation of many-body quantum systems and early-fault tolerant quantum computing
Apply here:
lanl.jobs/search/jobde...
Reposts appreciated!
Apply here:
lanl.jobs/search/jobde...
Reposts appreciated!
Quantum Computing and Early Fault-Tolerant Simulations at Los Alamos National Laboratory
Los Alamos National Laboratory is Hiring! Search available jobs or submit your resume now by visiting this link. Please share with anyone you feel would be a great fit.
lanl.jobs
October 22, 2025 at 10:22 PM
Our group is looking to hire postdocs to work at the intersection of quantum computing, quantum algorithms, quantum machine learning, simulation of many-body quantum systems and early-fault tolerant quantum computing
Apply here:
lanl.jobs/search/jobde...
Reposts appreciated!
Apply here:
lanl.jobs/search/jobde...
Reposts appreciated!
Reposted by Zoe Holmes
A new article introduces a continuous-variable analogue of the Pauli propagation algorithm, called Displacement Propagation. Surprisingly, non-Gaussianity and symplectic coherence can make the system easier to simulate when noise is present.
arxiv.org/abs/2510.07264
arxiv.org/abs/2510.07264
When quantum resources backfire: Non-gaussianity and symplectic coherence in noisy bosonic circuits
Analyzing the impact of noise is of fundamental importance to understand the advantages provided by quantum systems. While the classical simulability of noisy discrete-variable systems is increasingly...
arxiv.org
October 9, 2025 at 8:41 PM
A new article introduces a continuous-variable analogue of the Pauli propagation algorithm, called Displacement Propagation. Surprisingly, non-Gaussianity and symplectic coherence can make the system easier to simulate when noise is present.
arxiv.org/abs/2510.07264
arxiv.org/abs/2510.07264
Among todays mad arxiv overflow is our preprint:
"When quantum resources backfire: Non-gaussianity and symplectic coherence in noisy bosonic circuits"
We introduce a path propagation classical simulation alg for bosonic circuits
And find a funky interplay between quantum resources and noise
🧵👇
"When quantum resources backfire: Non-gaussianity and symplectic coherence in noisy bosonic circuits"
We introduce a path propagation classical simulation alg for bosonic circuits
And find a funky interplay between quantum resources and noise
🧵👇
October 9, 2025 at 11:48 AM
Among todays mad arxiv overflow is our preprint:
"When quantum resources backfire: Non-gaussianity and symplectic coherence in noisy bosonic circuits"
We introduce a path propagation classical simulation alg for bosonic circuits
And find a funky interplay between quantum resources and noise
🧵👇
"When quantum resources backfire: Non-gaussianity and symplectic coherence in noisy bosonic circuits"
We introduce a path propagation classical simulation alg for bosonic circuits
And find a funky interplay between quantum resources and noise
🧵👇
Reposted by Zoe Holmes
Quantum arXiv today.
October 9, 2025 at 8:20 AM
Quantum arXiv today.
Oh yeah. Strange. It was working (for me) earlier but not now.
Thanks for highlighting.
Let‘s try this: arxiv.org/abs/2510.01154
Thanks for highlighting.
Let‘s try this: arxiv.org/abs/2510.01154
Advantage for Discrete Variational Quantum Algorithms in Circuit Recompilation
The relative power of quantum algorithms, using an adaptive access to quantum devices, versus classical post-processing methods that rely only on an initial quantum data set, remains the subject of ac...
arxiv.org
October 2, 2025 at 5:50 PM
Oh yeah. Strange. It was working (for me) earlier but not now.
Thanks for highlighting.
Let‘s try this: arxiv.org/abs/2510.01154
Thanks for highlighting.
Let‘s try this: arxiv.org/abs/2510.01154
This is the paper: scirate.com/arxiv/2510.0...
Thanks for the fun collaboration Sasha (
@sheffield-qc.bsky.social
) and Chiddy !
Thanks for the fun collaboration Sasha (
@sheffield-qc.bsky.social
) and Chiddy !
scirate.com
October 2, 2025 at 1:27 PM
This is the paper: scirate.com/arxiv/2510.0...
Thanks for the fun collaboration Sasha (
@sheffield-qc.bsky.social
) and Chiddy !
Thanks for the fun collaboration Sasha (
@sheffield-qc.bsky.social
) and Chiddy !
FYI, our results here don’t contradict arXiv:2312.09121 which focus on loss estimation, proofs, and continuous protocols
Of these, the most intriguing/significant is the switch to discrete optimization
& maybe the path to adaptive quantum advantage is all about finding those discrete sweet spots 😉
Of these, the most intriguing/significant is the switch to discrete optimization
& maybe the path to adaptive quantum advantage is all about finding those discrete sweet spots 😉
October 2, 2025 at 1:23 PM
FYI, our results here don’t contradict arXiv:2312.09121 which focus on loss estimation, proofs, and continuous protocols
Of these, the most intriguing/significant is the switch to discrete optimization
& maybe the path to adaptive quantum advantage is all about finding those discrete sweet spots 😉
Of these, the most intriguing/significant is the switch to discrete optimization
& maybe the path to adaptive quantum advantage is all about finding those discrete sweet spots 😉
And we have found one… we provide numerical evidence that our problem lives in the Goldilocks zone:
- trainable (no exponential concentration)
- not classically surrogatable (thanks to high entanglement + magic)
- trainable (no exponential concentration)
- not classically surrogatable (thanks to high entanglement + magic)
October 2, 2025 at 1:23 PM
And we have found one… we provide numerical evidence that our problem lives in the Goldilocks zone:
- trainable (no exponential concentration)
- not classically surrogatable (thanks to high entanglement + magic)
- trainable (no exponential concentration)
- not classically surrogatable (thanks to high entanglement + magic)
Of course, this being quantum, we face two extra demons…
- Exponential concentration (barren plateaus + shot noise)
- Classical surrogation (can classical shadows fake the landscape?)
For a real separation, we need a sweet spot that dodges both.
- Exponential concentration (barren plateaus + shot noise)
- Classical surrogation (can classical shadows fake the landscape?)
For a real separation, we need a sweet spot that dodges both.
October 2, 2025 at 1:23 PM
Of course, this being quantum, we face two extra demons…
- Exponential concentration (barren plateaus + shot noise)
- Classical surrogation (can classical shadows fake the landscape?)
For a real separation, we need a sweet spot that dodges both.
- Exponential concentration (barren plateaus + shot noise)
- Classical surrogation (can classical shadows fake the landscape?)
For a real separation, we need a sweet spot that dodges both.
More concretely, we show that for a range of moderate entangling strengths the landscape is unimodal but non-separable landscapes.
Then numerics show adaptive hill-climbing converges efficiently
But non-adaptive approaches - blow up exponentially.
Then numerics show adaptive hill-climbing converges efficiently
But non-adaptive approaches - blow up exponentially.
October 2, 2025 at 1:23 PM
More concretely, we show that for a range of moderate entangling strengths the landscape is unimodal but non-separable landscapes.
Then numerics show adaptive hill-climbing converges efficiently
But non-adaptive approaches - blow up exponentially.
Then numerics show adaptive hill-climbing converges efficiently
But non-adaptive approaches - blow up exponentially.
We translate that logic into a quantum recompilation task
The hidden string = the placement of T-gates between layers of semi-random unitaries
Goal = uncover the T gates positions
As in LeadingOnes, identifying early T-gates helps you make progress, but you can’t optimize each gate independently
The hidden string = the placement of T-gates between layers of semi-random unitaries
Goal = uncover the T gates positions
As in LeadingOnes, identifying early T-gates helps you make progress, but you can’t optimize each gate independently
October 2, 2025 at 1:23 PM
We translate that logic into a quantum recompilation task
The hidden string = the placement of T-gates between layers of semi-random unitaries
Goal = uncover the T gates positions
As in LeadingOnes, identifying early T-gates helps you make progress, but you can’t optimize each gate independently
The hidden string = the placement of T-gates between layers of semi-random unitaries
Goal = uncover the T gates positions
As in LeadingOnes, identifying early T-gates helps you make progress, but you can’t optimize each gate independently
This is the canonical “adaptivity pays” task
Its unimodal (no local minima) but non-separable (each bit cannot be trained independently)
- Adaptive strategies can flip one bit at a time, use the feedback, and find the string in O(n) queries.
- Non-adaptive strategies need exponentially many.
Its unimodal (no local minima) but non-separable (each bit cannot be trained independently)
- Adaptive strategies can flip one bit at a time, use the feedback, and find the string in O(n) queries.
- Non-adaptive strategies need exponentially many.
October 2, 2025 at 1:23 PM
This is the canonical “adaptivity pays” task
Its unimodal (no local minima) but non-separable (each bit cannot be trained independently)
- Adaptive strategies can flip one bit at a time, use the feedback, and find the string in O(n) queries.
- Non-adaptive strategies need exponentially many.
Its unimodal (no local minima) but non-separable (each bit cannot be trained independently)
- Adaptive strategies can flip one bit at a time, use the feedback, and find the string in O(n) queries.
- Non-adaptive strategies need exponentially many.
Our task is a quantum twist on the classic LeadingOnes-OneMax problem.
In this problem you're trying to learn a hidden bitstring.
Your score = how many leading bits match the target before the first mismatch.
So 1110… matches 1101 better than 1011… even if they have the same Hamming weight.
In this problem you're trying to learn a hidden bitstring.
Your score = how many leading bits match the target before the first mismatch.
So 1110… matches 1101 better than 1011… even if they have the same Hamming weight.
October 2, 2025 at 1:23 PM
Our task is a quantum twist on the classic LeadingOnes-OneMax problem.
In this problem you're trying to learn a hidden bitstring.
Your score = how many leading bits match the target before the first mismatch.
So 1110… matches 1101 better than 1011… even if they have the same Hamming weight.
In this problem you're trying to learn a hidden bitstring.
Your score = how many leading bits match the target before the first mismatch.
So 1110… matches 1101 better than 1011… even if they have the same Hamming weight.
We provide evidence of an exponential gap between adaptive & nonadaptive strategies for a quantum recompilation task
Key takeaways:
- Entanglement isn’t always a roadblock: its degree can aid training
- Discrete optimization may be key to finding sweet spots between concentration & surrogation
Key takeaways:
- Entanglement isn’t always a roadblock: its degree can aid training
- Discrete optimization may be key to finding sweet spots between concentration & surrogation
October 2, 2025 at 1:23 PM
We provide evidence of an exponential gap between adaptive & nonadaptive strategies for a quantum recompilation task
Key takeaways:
- Entanglement isn’t always a roadblock: its degree can aid training
- Discrete optimization may be key to finding sweet spots between concentration & surrogation
Key takeaways:
- Entanglement isn’t always a roadblock: its degree can aid training
- Discrete optimization may be key to finding sweet spots between concentration & surrogation