lenz3000.bsky.social
@lenz3000.bsky.social
Whoop whoop, this got accepted at Neurips 25 🎉🎉🎉
I'll make an update once we have the updated version
I'll make an update once we have the updated version
September 18, 2025 at 3:46 PM
Whoop whoop, this got accepted at Neurips 25 🎉🎉🎉
I'll make an update once we have the updated version
I'll make an update once we have the updated version
If you care about improving the performance of Flow Matching, especially in molecular modeling, this might be for you.
No redesign. No retraining. Just better gradients.
arxiv.org/abs/2505.10139
By @leonklein.bsky.social & me.
No redesign. No retraining. Just better gradients.
arxiv.org/abs/2505.10139
By @leonklein.bsky.social & me.
Path Gradients after Flow Matching
Boltzmann Generators have emerged as a promising machine learning tool for generating samples from equilibrium distributions of molecular systems using Normalizing Flows and importance weighting. Rece...
arxiv.org
May 19, 2025 at 9:25 AM
If you care about improving the performance of Flow Matching, especially in molecular modeling, this might be for you.
No redesign. No retraining. Just better gradients.
arxiv.org/abs/2505.10139
By @leonklein.bsky.social & me.
No redesign. No retraining. Just better gradients.
arxiv.org/abs/2505.10139
By @leonklein.bsky.social & me.
The result?
Replicating the experiments from Equivariant Flow Matching (2306.15030), using the same limited resources, we get:
✓ up to 3× higher ESS on LJ55
✓ 2x on Alanine Dipeptide XTB
✓ Improved performance on most tasks
Replicating the experiments from Equivariant Flow Matching (2306.15030), using the same limited resources, we get:
✓ up to 3× higher ESS on LJ55
✓ 2x on Alanine Dipeptide XTB
✓ Improved performance on most tasks
May 19, 2025 at 9:23 AM
The result?
Replicating the experiments from Equivariant Flow Matching (2306.15030), using the same limited resources, we get:
✓ up to 3× higher ESS on LJ55
✓ 2x on Alanine Dipeptide XTB
✓ Improved performance on most tasks
Replicating the experiments from Equivariant Flow Matching (2306.15030), using the same limited resources, we get:
✓ up to 3× higher ESS on LJ55
✓ 2x on Alanine Dipeptide XTB
✓ Improved performance on most tasks
Path Gradients were the topic of my dissertation — and they turn out to be a great fine-tuning step for Boltzmann Generators.
Just apply them after Flow Matching.
Fine-tuning reliably improves performance without big changes to the overall learned distribution.
For example, on alanine dipeptide:
Just apply them after Flow Matching.
Fine-tuning reliably improves performance without big changes to the overall learned distribution.
For example, on alanine dipeptide:
May 19, 2025 at 9:17 AM
Path Gradients were the topic of my dissertation — and they turn out to be a great fine-tuning step for Boltzmann Generators.
Just apply them after Flow Matching.
Fine-tuning reliably improves performance without big changes to the overall learned distribution.
For example, on alanine dipeptide:
Just apply them after Flow Matching.
Fine-tuning reliably improves performance without big changes to the overall learned distribution.
For example, on alanine dipeptide: