iiinova — where your digital self keeps evolving

Mark Mitchison @mitchison.bsky.social · 25/12/2025

Merry Xmas BlueSky!! Breakfast today is chicken mappas and appam… They know how to celebrate in Kerala 🤩🎄

Chicken mappas (a kind of curry) and appam (a fluffy rice-based savoury pancake)

170

Mark Mitchison @mitchison.bsky.social · 17/11/2025

Here’s the whole crew at the film’s world premiere at AMBER research centre last week 🥰

A picture of 9 people who contributed to making Quantum Quest, standing in front of a large poster with AMBER written on it, and including the cartoon characters Alice, Bob and Cat.

030

Mark Mitchison @mitchison.bsky.social · 16/11/2025

But can a clock—which distinguishes past from future—be truly reversible? Of course not! The resolution is that monitoring equilibrium fluctuations requires a detector that is out of equilibrium, producing entropy. Our friends @naresgroup.bsky.social showed this in a proof-of-principle experiment.

A picture showing a quantum clock (left) driven by a thermal gradient, whose ticks are transitions between microscopic quantum states. Reading out these transitions requires an active tick sensor, which produces a large amount of entropy per tick of the clock.

131

Mark Mitchison @mitchison.bsky.social · 18/10/2025

We find strong system-bath coupling can suppress current fluctuations in a driven quantum system, allowing violation of thermodynamic uncertainty relations! We understand this in terms of the system hybridising with a collective environment mode, which leads to antibunching in the current statistics

Diagram of our system: a driven qubit coupled to a collective environment mode (the reaction coordinate).

000

Mark Mitchison @mitchison.bsky.social · 01/10/2025

Awesome to have Kavan Modi visit us at King’s and talk about his work characterising correlated errors in quantum computers, using Quantum Process Tensor Tomography 🤩

Kavan Modi standing in front of a projector slide titled Quantum Process Tomography

040

Mark Mitchison @mitchison.bsky.social · 26/08/2025

So many great talks at Fundamental Limits of Quantum Technologies already, and many more yet to come! Here’s Shane Dooley of DIAS telling us about metrology using many-body quantum systems 🤩 www.aspects-quantum.com/flqt2025 @aspects-quantum.bsky.social

Dr Shane Dooley standing on the stage, next to the first slide of his talk titled “Remote quantum sensing through a thermal Floquet spin chain”

081

Mark Mitchison @mitchison.bsky.social · 26/08/2025

Fresh on arXiv! arxiv.org/abs/2508.16375 By casting detectors as autonomous thermal machines, we uncover the link between entropy production and the detector’s efficiency, timing jitter, dead time and dark counts. Outstanding work by Manu Schwarzhans with my @aspects-quantum.bsky.social friends ❤️

Illustration of the three-way trade off between dark count rate (top), detection time uncertainty (jitter, bottom left) and dead time (bottom right). These cannot all be simultaneously improved, but must be traded off against one another, if the detector’s rate of steady-state entropy production (centre) is fixed.

061

Mark Mitchison @mitchison.bsky.social · 17/07/2025

I very much appreciate a well organised metro map (Singapore MRT)

020

Mark Mitchison @mitchison.bsky.social · 08/07/2025

Very proud and honoured to kick off QTD Singapore 2025 with an invited talk! I attended my first QTD conference in Porquerolles 10 years ago: a career-changing moment where I found my community and cemented my love for the topic 🤓 I hope some of this year’s attendees get as much out of it as I did!

Mark on stage giving a talk at QTD 2025. The slide motivates why the audience of quantum thermodynamics researchers should be interested in clocks.

0120

Mark Mitchison @mitchison.bsky.social · 16/04/2025

How to compute heat statistics at strong coupling? We introduce a heat operator, casting two-point measurement statistics as a simple expectation value. This makes it easy to tackle using tensor network algorithms eg. the chain mapping. Many exciting applications for this one! 🤩

A diagram showing a tensor network state, evolved in time by a unitary operator, followed by an expectation value of the heat operator (raised to the nth power). This yields the nth statistical moment of the heat distribution.

100

Mark Mitchison @mitchison.bsky.social · 14/04/2025

Happy World Quantum Day everyone!

0110

Mark Mitchison @mitchison.bsky.social · 01/04/2025

Today I start a new job at King’s College London! I’m now a Proleptic Senior Lecturer and Royal Society University Research Fellow in the Dept. of Physics @kingsnmes.bsky.social I’m delighted to return home after >8 years working abroad, and very excited to start this new chapter in my career!

Selfie of me standing in front of the entrance to King’s College London Strand campus

1201

Mark Mitchison @mitchison.bsky.social · 26/03/2025

Extremely proud to be holding a beautifully bound copy of the thesis of Oisín Culhane, my first PhD student!

Picture of me holding a thesis titled “Thermodynamics of Nanoelectromechanics”

060

Mark Mitchison @mitchison.bsky.social · 20/03/2025

Pauli Erker giving a wonderful invited talk on the autonomous quantum processing unit arxiv.org/abs/2402.00111, part of an inspiring session on Time in Quantum Mechanics and Thermodynamics at #APSSummit2025

061

Mark Mitchison @mitchison.bsky.social · 19/03/2025

At #APS watching Simon Sundelin nailing his talk on our @aspects-quantum.bsky.social collaboration 😍

Simon Sundelin speaking at a podium at APS

One of Simon’s slides explaining the experimental setup

041

Mark Mitchison @mitchison.bsky.social · 22/02/2025

These differences appear to persist in the thermodynamic limit: finite-size scaling shows the entanglement variance for falls off exponentially only in chaotic models. This suggests that basically any local multi-time intervention on a large chaotic many-body system generates maximal entanglement

110

Mark Mitchison @mitchison.bsky.social · 22/02/2025

Not only does this recover several previously studied chaos and complexity measures, it contains a bunch of new information in its higher moments! Chaotic many-body systems generate a distribution closely matching the Haar-random process, unlike integrally/localised models

220

Mark Mitchison @mitchison.bsky.social · 22/02/2025

But chaos is identified by the complexity of the output states as well as the time series. We capture this with the projected process ensemble: an ensemble of states conditioned on multi-time interventions. This is like the projected ensemble studied in deep thermalisation, but for quantum processes

120

Mark Mitchison @mitchison.bsky.social · 22/02/2025

Quantum chaos does not have a unique definition. One approach in classical physics is to think about the complexity of the trajectories in phase space over time — the dynamical entropy. Operationally, a trajectory of a quantum system is a sequence of measurements, described by a process tensor

110

Mark Mitchison @mitchison.bsky.social · 04/02/2025

Our paper shows this won't work! Any quantum clock must produce macroscopic amounts of entropy, in order to amplify its ticks into a classical record of time. The entropy produced by the measurement dwarfs that needed to generate ticks, and this dissipation can even drive the clock by itself 4/6

Plots of the clock's precision as a function of entropy produced by generating ticks (left) and entropy produced by measuring and amplifying the ticks (right). The latter is some 9-10 orders of magnitude larger than the former.

110

Mark Mitchison @mitchison.bsky.social · 04/02/2025

Latest preprint from the ASPECTS consortium! arxiv.org/abs/2502.00096 Experiments by the amazing @naresgroup.bsky.social and theory by @mathsmire.bsky.social of @quitphysics.info We measured the entropy cost of generating and recording ticks in a microscopic clock 1/6 🧵👇

Depiction of the quantum-classical transition in a microscopic clock. A quantum clock transitions between three discrete states, driven by thermal noise. These transitions are amplified by a sensor into a classical signal, which can be shared among classical observers as an objective record of the passage of time.

184

Mark Mitchison @mitchison.bsky.social · 25/01/2025

I’m probably late to this party, but can we please talk about APS butchering their own website? Look at the formatting in PRL’s author “style guide” for example… are those supposed to be section headings?! Link previews in Slack no longer work either. Why did they do this?!

030

Mark Mitchison @mitchison.bsky.social · 19/12/2024

A proud moment as my first PhD student, Oisín Culhane, defends his excellent thesis on the quantum and stochastic thermodynamics of tiny wobbly things 😍

080

Mark Mitchison @mitchison.bsky.social · 03/12/2024

Giving a seminar in the Michael Berry Lecture Theatre @bristoluni.bsky.social! ❤️🫐

040

Mark Mitchison @mitchison.bsky.social · 02/12/2024

🚨new paper🚨 out today! We use the same kind of artificial intelligence model that underpins ChatGPT for quantum feedback control! The “attention” mechanism that lets generative AI understand context also helps us deal with memory effects in the measurement record! arxiv.org/abs/2411.19253

Diagram of our new neural network architecture for measurement based quantum feedback control

171

Mark Mitchison @mitchison.bsky.social · 02/12/2024

Ok BlueSky let’s do this! I’m a theoretical physicist interested in how very tiny ⚛️ or very cold 🥶 bits of matter interact with their surroundings: physicists call these *open quantum systems*. I’m going to be posting on my research and public outreach activities, plus the odd bit of fun/nonsense 🤪

A picture of me standing in front of a modern abstract painting

0192