James Alvey

Postdoc @GRAPPAInstitute | Cosmology ∩ ML ∩ Particle Physics

[Benjamin Wandelt @bwandelt]

Time-series forecasts drive countless decisions with vast, global economic impact.

Our paper "Scaling-laws for Large Time-series Models" shows that prediction performance of Large Time-series Models obeys scaling laws, just like Large Language Models!

https://t.co/WLMTFyMTTQ

[Benjamin Wandelt @bwandelt]

Kudos to lead author Thomas Edwards and the amazing team: James Alvey, Justin Alsing & Nam Nguyen
@JohnsHopkins @FlatironCCA @astroIAP @GRAPPAinstitute

Check out our paper for the details!
https://t.co/WLMTFyMTTQ

[Samaya Nissanke (she/her) 🕊️❤️ @samayanissanke]

New group & @GRAPPAinstitute paper: many congratulations & so proud of group PhD student @Uddi9, grappa postdoc @James_A_42 first machine learning PEREGRINE paper on gravitational wave parameter inference with @ai4science_lab @bkmi13, @C_Weniger https://t.co/QNt2ssxc0f

[GRAPPA Centre @GRAPPAinstitute]

Check out this amazing new GRAPPA paper led by PhD student @Uddi9 and postdoc @James_A_42 ! The authors introduce PEREGRINE, a new code that can perform simulation-based inference on a broad class of gravitational wave signals

[GRAPPA Centre @GRAPPAinstitute]

Another super interesting paper https://t.co/Z8NKpCHxc6 led by @GRAPPAinstitute postdoc @James_A_42 and PhD student @mathisgerdes . The authors introduce ALBATROSS, a new simulation-based inference code, this time designed to analyze the complicated stellar streams!

[James Alvey @James_A_42]

📦 Takeaway: We found that even when allowing for a lot of variation in the concrete model, it really is hard for DM to hide in this window!

Thread was a bit longer than planned, but feel free to ask any questions!
P.S. Code will soon be released here: https://t.co/SLVd5iOVbi

[James Alvey @James_A_42]

Is there anywhere for DM to hide❓💫
Or, what happens when cosmic rays bump into DM❓

Result of a really enjoyable collaboration with Torsten Bringmann and Helena Kolešová, started during a visit to @CERN last year!

Link to paper: https://t.co/KEGR5zotuO
(Long-ish 🧵 below)

[James Alvey @James_A_42]

🧩 Summary: Overall, the idea of CRDM is not a new one, but it is also an important concept in the regime of strongly interacting DM. So, we dedicated some time to better modelling of the various aspects of the process, specifically the inelastic scattering in the rock.

[James Alvey @James_A_42]

Congrats Dr. Sabti @NashIn0044, don’t know anyone who deserves it more. Very happy to be able to celebrate with you in NL 🇳🇱