It’s been three weeks since I concluded my term as Exotics convener of the ATLAS collaboration, a role which put me at the heart of the programme of searches for new particles and interactions at the LHC fo two years. It’s been an intense but fulfilling ride and I wanted to dedicate a post to reviewing some of what I achieved and a few of the lessons I’ve learned.
A tiny gesture can make a huge difference: building a sense of community
This first achievement was more sociology than science. During my term, I focused on strengthening the sense of community in the Exotics group. There are about 550 physicists in the Exotics group, working on a vast array of different topics, from long-lived particles, to heavy neutral leptons, leptoquarks, vectorlike-leptons, quantum black holes and heavy and light resonances. But ultimately, we’re all interested in the same questions: the nature of dark matter, the matter–antimatter asymmetry of the universe, the hierarchy problem, the origin of neutrino masses… It’s just that we can’t all do everything, so people have to choose what to dedicate their time to. Nevertheless, the successes of the Exotics group are shared successes. If we make a big advance in one topic, it matters to the people working on other problems too.
That’s why one thing I am particularly proud of, small as it may seem, was that I fostered a culture where we publicly celebrate the successes of other community members. For example, we created a discussion channel for people to post pictures from conferences where talks showed our recent results on the big screen, and celebrate our speakers.
It has become very popular, and it is a great pleasure to keep an eye on during the conference season. If you’ve spent three years working on a result, it’s so gratifying to see a photo of one’s hard-earned achievement being discussed by scientists in big meetings. In another example, I made a point every Friday evening, in our community forum, to give a shout out to all the teams who crossed important milestones that week : appointments to coordination roles, formation of new analyses, unblinding requests, successful approvals, circulation to the collaboration, arxiv submission, publication, etc… The purpose was to remind people that their work and their successes are valued and shared, and that there is a lot going on in our community! In my digital farewell card, a lot of people commented on how much they appreciated these channels and the sense of shared celebration, and I hope they remain part of the Exotics culture. It may seem trivial compared to the physics, but one thing I have learned is that people who feel valued, seen and motivated, can do their best work. Perhaps it’s no coincidence that during my term, we got a whole load of analyses, which had beed “stuck” for months or years, into review, and out into the public domain.
Lead from the front, and people will follow: a focus on analysis preservation
A more scientifically-substantial undertaking was to put an important emphasis on analysis preservation and re-useability of our work during my mandate. Analysis preservation has been a pet projects of mine for years: the gist is that we should publish enough material publicly for our results to be reusable well into the future. It’s trickier than it sounds. But this is especially important because the next searches to come out might be the last for a very long time: with more data and more challenging conditions at the high-luminosity LHC, sensitivity will take a while to catch up to where we are now. We need to make sure that the results we publish now are still useable for years, or even decades.
As a result of various policies I put in place, the Exotics group achieved 100% coverage on the “HEPData” portal (which stores digitised plot information as well as crucial additional analysis preservation content) for 2025 papers, and we are on track for the same for 2026 papers. The last time we had 100% coverage was in 2020, for a smaller number of papers. If we manage it for 2026, it will be the first time in Exotics history that two consecutive years have had full coverage. My aim has been to make that sustainable, so we can continue to hit 100% indefinitely from now on, progressively adding richer and richer content to our records.
We achieved this by building a tool to streamline the creation of the HEPData entries, removing many of the common time-sinks. This was also helped by the timely arrival of AI models capable of helping with coding, which could take that tool and build records under human supervision, and with a derailed human validation. But ultimately, I also had to take over myself the job for teams who were exhausted and unable/unwilling to do it themselves (in the end, I built about half of the entries myself, and about half of those with AI assistance). At first this seemed like a sort of failure, but then I realised that having one expert HEPData record builder makes for more consistent, easier-to-validate records (since common problems have already been encountered, and there is no learning curve), significantly lowering the human time needed for the task, with considerably better outcomes. This also had the benefit of leaving more room to focus on the quality of the entries rather than just their existence, because if all a team has to do is provide the numerical values, I can ask them for a richer or more varied materials without exhausting them.
I was already proud enough that we managed to get to that record, but I soon discovered an additional bonus. For the 2026 papers, I expected to have to build many of the entries myself again. But when I asked the teams, I found that many replied, “Don’t worry, we are already on it!”: something that I had never heard before (the usual response used to be “Urg…do we really have to…?”). I believe that by achieving initial 100% coverage milestone, casting it as a community success, and projecting that making the HEPData records was valued by the leadership (if nothing else because the leader of your community –someone with practically zero spare time and a zillion things to do– is prepared to do it himself if needed), we motivated teams to go the extra mile. And that is how I learned a lesson which I think will stick with me for many years: if you lead from the front, people will follow.
Don’t just turn the handle: building a strategy for the future
I last big achievement I want to discuss relates to strategy. Because of the timing of my mandate, I saw the start of the LHC Run 3 analysis campaign: the first Exotics search using part of that dataset came out under my watch. One of my aims was also to wrap up the Run 2 campaign, which I didn’t quite manage: there are two remaining searches which did not make it into the review pipeline during my term, but they had complicating factors outside of anyone’s control.
However, beyond just turning the handle and keeping the show on the road, something very important to me was to think carefully about how to use the resources at our disposal to best plan for the future of the programme. As I’ve said before in blog posts, seminars, conferences and internal meetings, “leave no stone unturned” is not a research strategy. It is an empty phrase, and an unrealistic one, because it assumes we have unlimited resources and time. What we need instead is a plan based on our priorities. Even understanding our priorities is a huge amount of work: it means understanding what we’ve done so far, the context of our research, and where it might lead. It’s much easier to just “turn the handle”, in other words, consider business as usual, reacting to opportunities and scientific trends as they come, leading to a piecemeal, scattershot programme fo research. But I wanted to do better than that.
That’s why, during my term, I put together a small team of experts for each of the Exotics subgroups, and tasked them to produce short “roadmap” documents. Each document sets out the history of the group, lists the key target models searched for, and proposes potential targets for the future. Those documents were prepared over the course of last year, and version two, which incorporated feedback from the wider community, was published internally about a week before the end of my term. As well as an awesome primer for newcomers, this documentation now encapsulates the key final states and models targeted by each group, and their motivation, while relating how we got to the “state of the art” of today’s experimental situation.
This was an important piece of work, which consolidated nebulous institutional knowledge, but it was not enough. A map should of course tell you were you are and how you got there, but it should also tell you where you are going. So there was a further challenge: to define specific targets for each subgroup for the Run 3 campaign. The effort was divided into two steps. The first was a set of 10 global objectives, with quantified goals, which we would like to achieve (and in some cases absolutely must achieve) by the end of the Run 3 data analysis campaign. We made these deliberately aggressive, to push beyond what we could get from just adding extra data alone: to achieve them, our community will need creativity and new techniques. As a second step, alongside these global targets, each subgroup defined a list of its own specific targets, more loosely scoped, to highlight the breadth of the programme and our sensitivity. These were developed with the help of our subgroup conveners, the current leaders of each sub-community.
The targets were presented to the Exotics group by Giuliano and I at a special meeting, just a week before I finished my term, which was a wonderful way to conclude my mandate. There was some very useful discussion afterwards, with further discussion planned at the Exotics workshop next week in Clermont-Ferrand. I consider this probably my most important contribution, as I hope it will give us the tools we need to truly make the most of the data we have at the LHC. The leadership lesson here? Although it can be easy to get caught in the day to day of meetings, paper reviews, and admin, one must make time to reflect on the bigger picture.
A few numbers
And to finish, here are a few stats from my term, taken from my farewell message to the Exotics community.
Together we got 27 papers out of the ATLAS collaboration and into the public domain, and dozens of summary plot PUB notes.
I did a few back of the envelope calculations.
Five subgroups, each with two hours meetings each week, for about 40 weeks a year over two years. Five two-to-two chats with subgroup conveners, roughly every other week on average, for 40 weeks. 27 papers out means, 27 Paper Approval Meetings and 27 Exotics approvals, each two hours long. Physics coordination meetings, 2h apiece, for 80 weeks. Ad hoc meetings, Exotics plenaries, etc… something like 40 or 50 hours. By my count, that’s around 1200 hours spent on Zoom.
27 papers, also means reading 27 paper drafts, on average 20 pages each. It means 27 internal notes, on average 150 pages each. Even if we account for the fact that we typically split this work between conveners, that’s over 2000 A4 pages of material. That’s about the same as reading À la Recherche Du Temps Perdu by Marcel Proust, often cited as the longest book ever written, twice over.
Handing over
And with that, on the 1st of October, I handed over the keys to the Exotics bandwagon, and I signed off to the community with one last shout-out. I am proud of what I achieved during that time. In addition to the above, I felt I ran the show effectively and efficiently, making sure all our tasks, big and small, where dealt with quickly. I made plenty of mistakes during my time, and there are certainly things I would have done differently, but three weeks after the dust has settled, I look back with pride at my mandate.
I leave you with this (AI-generated) image which my co-convener Giuliano cooked up, to visualise how we thinks my life is now that my mandate is over.
But of course, the next challenge always comes along fast! I was asked to take over as one of the ATLAS representatives to the LHC “Beyond the Standard Model” working group, which unites researches from ATLAS, CMS, LHCb, Theory and beyond, to coordinate on important topics such as which benchmark models to use, making common summary plots, and open data policies. I started that on October 1st. Moreover, I was asked to stand as ATLAS Education and Outreach officer. The election was last week, and I am proud to say that the collaboration has placed its trust in me for that role, starting in March next year. Watch this space for a future blog article about why I think this is a particularly critical time to be taking on such a role, and what I have planned for it.
AI disclaimer: The initial draft for this article was dictated to my phone while I was travelling to CERN this week. I used Claude to tidy up the transcription into something readable and grammatically correct, and then used that as a base for the article, which I extended and reworked. All mistakes, typos, repetitions and opinions are my own!