✨ New preprint!
LLMs can generate text & images — but can they also discover equations in high energy physics?
Using the LLM-SR framework, we show they can uncover compact, interpretable forms for collider observables at the LHC!
More details in:
🔗 arxiv.org/abs/2510.03397
Great news! Gauge Theory of Elementary Particle Physics by Ta-Pei Cheng and Ling-Fong Li (Oxford) is now OPEN ACCESS! 🎉 You can read and download it for free using the PDF link below. 📖✨
fdslive.oup.com/www.oup.com/...
The papers and posters for our Machine Learning and Physical Sciences workshop at #NeurIPS2024 are online #ml4ps2024. Come check it out on Sunday
ml4physicalsciences.github.io/2024/
Shoutout to @milescranmer.bsky.social's PySR, which made the task smoother! Open-source making the community stronger!
3/n: 🤖 SR combines analytic simplicity with the predictive capabilities of ML, providing a valuable tool for simplifying pheno analyses in colliders!
2/n: 🔬 QED provides a solid benchmark for validating SR against first principles. After this validation, we apply SR to Drell-Yan structure functions, deriving simple formulas as an alternative to methods like NNs (with many trainable parameters) or fixed functional forms
1/n: 🎯 How can machine learning simplify phenomenological analyses at the LHC? We explore symbolic regression (SR) as a tool to derive compact, precise analytic expressions. First stop: testing SR on quantum electrodynamics (QED).
The power of ML and the intuition of analytical expressions in collider simulations!
Happy to be at the @neuripsconf.bsky.social Machine Learning for the Physical Sciences Workshop presenting 'Symbolic Regression for Precision LHC Physics'!
ml4physicalsciences.github.io/2024/
Hi Ben, could I be added please?
