---
title: "Fudan's Plug-in Module Makes Protein AI Explore 37x Faster Without Retraining"
subtitle: "A new method injects a history-aware bias into frozen protein dynamics emulators, reaching rare conformational states up to 37x faster than unbiased rollouts"
url: "https://alphasignal.ai/news/fudan-s-plug-in-module-makes-protein-ai-explore-37x-faster-without-retraining"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-06-01T00:00:00.000Z"
updated: "2026-06-09T10:15:21.658Z"
type: "Paper"
topics:
  - "Llms"
  - "Training Infra"
subtopics:
  - "Hallucinations"
  - "Pretraining"
upvotes: 2117
source_url: "https://arxiv.org/abs/2606.01833"
comments: 0
truncated: true
---

# Fudan's Plug-in Module Makes Protein AI Explore 37x Faster Without Retraining

> A new method injects a history-aware bias into frozen protein dynamics emulators, reaching rare conformational states up to 37x faster than unbiased rollouts

Generative models have become a compelling shortcut for studying how proteins move. Instead of running expensive molecular dynamics (MD) simulations that can take days on specialized hardware, these **generative emulators** learn to produce plausible protein trajectories at a fraction of the cost. But they come with a fundamental flaw: they keep revisiting the same familiar shapes. A new [paper from Cheng et al.](https://arxiv.org/abs/2606.01833), a multi-institutional collaboration spanning Fudan University, the Shanghai Academy of AI for Science, and the Shanghai Innovation Institute, proposes a principled fix inspired by a decades-old trick from classical simulation.

## The Trap Every Emulator Falls Into

To understand the problem, it helps to know what these emulators are doing. A generative emulator is trained on short MD trajectories and learns to autoregressively generate new protein conformations one frame at a time, conditioned on a few preceding frames. This is far cheaper than physics-based simulation, but it inherits a critical bias: the model was trained on a particular distribution of structures, and when you run it for a long time, it gravitates back toward those known states.

The structures that actually matter for biology, things like the rare open state of an enzyme active site or a cryptic drug-binding pocket, sit behind high free-energy barriers. These rare states underlie biological processes including catalysis, transport, and folding, but sit behind high free-energy barriers that direct MD rarely crosses on routine timescales. An emulator trained on short trajectories almost never crosses those barriers either, so long-horizon rollouts just circle the starting structure.

MD itself struggles with a vast separation of timescales between femtosecond-level integration steps and millisecond-level transitions often required for full exploration of the conformational landscape. Classical enhanced-sampling methods like metadynamics and umbrella sampling were invented to solve exactly this problem for physics-based simulation, but they still require running the expensive simulator itself. The question this paper asks is: can you get the same exploration boost inside a generative model, without touching the model's weights?

## Metadynamics, but in Diffusion Space

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [Learning Implicit Bias in Generative Spaces for Accelerating Protein Dynamics Emulation](https://arxiv.org/abs/2606.01833)
- [AI-based Methods for Simulating, Sampling, and Predicting Protein Ensembles](https://arxiv.org/html/2509.17224v1)
- [Scalable emulation of protein equilibrium ensembles with generative deep learning (BioEmu)](https://www.science.org/doi/10.1126/science.adv9817)
- [Accelerated sampling of protein dynamics using BioEmu augmented molecular simulation](https://www.biorxiv.org/content/10.64898/2026.01.07.698041.full.pdf)