Bert de Vries @bertdv0

Trying to apply the free energy principle to engineering problems, where 'trying' means: minimizing free energy. bertdv.nl Eindhoven, Netherlands Joined January 2014

Tweets

213
Followers

530
Following

674
Likes

503

Wouter Nuijten @wouterwln

3 weeks ago

Out now in TMLR! 🚀 We formalize how epistemic priors transform Expected Free Energy minimization into a standard variational objective. This allows us to frame planning as a continuous variational optimization problem, moving away from combinatorial tree search. 👇

1 1 2 130 0

View Details

Bert de Vries @bertdv0

4 months ago

The result: a principled framework for embodied, agentic AI — robots, drones, and autonomous systems that perceive, learn, and act on-the-fly in real time.

0 0 0 75 0

View Details

Bert de Vries @bertdv0

4 months ago

Finally: Active Inference (AIF). AIF extends BML to embodied agents with a full commitment to variational inference for state estimation, learning, planning, and control. See t.ly/zcSdH

1 0 0 93 0

View Details

Bert de Vries @bertdv0

4 months ago

I recently had the pleasure to lecture at the Machine Learning Summer School in Melbourne t.ly/hX0Uo on Bayesian Machine Learning → Active Inference. All materials (slides + notebooks) available at github.com/bertdv/mlss-20… . Thread below 👇

1 2 7 524 2

View Details

Wouter Nuijten @wouterwln

10 months ago

This is what we ask ourselves in our latest paper (arxiv.org/abs/2508.02197), where we solve Minigrid environments with Active Inference. (Spoiler, it works!) Code available at github.com/biaslab/EFEasV… Done together with @mlmykola , @ThijsvdLaar and @bertdv0

1 2 5 284 2

View Details

Wouter Nuijten @wouterwln

10 months ago

What's even nicer: because our method injects priors locally, everything still works within @ReactiveBayes ' RxInfer.jl using message passing. Special thanks to my colleagues at @LazyDynamics for making this possible!

0 1 4 219 1

View Details

Wouter Nuijten @wouterwln

10 months ago

In Active Inference, a lot of time is spent on computing Expected Free Energy. What if we could tweak the generative model such that EFE can be minimised with traditional variational inference methods?

1 3 6 390 2

View Details

Lazy Dynamics @LazyDynamics

12 months ago

Backed Trojan Robotics (Team 24090) at the FIRST® Tech Challenge European Premier Event in Eindhoven (July 1–5)  . They hustled—coding, building, troubleshooting—and came away with 3rd in the Think Award. Proud to support their next steps. 🚀 #FTC #Robotics #STEM

0 1 2 312 0

View Details

Bert de Vries @bertdv0

a year ago

@MovingFramesP @fchollet Technically, FEP leads to simultaneously active inference (updating states), active learning and active model selection. The term “active inference” is usually interpreted as an umbrella over all active free energy minimizing processes.

0 0 2 35 0

View Details

Bert de Vries @bertdv0

a year ago

@fchollet x.com/bertdv0/status…

Bert de Vries @bertdv0

a year ago

Agreed with @fchollet on FEP (t.ly/s4gMV), but FEP is more than a pretty good idea, and there are more benefits to realizing an agent as an active inference (AIF) process beyond active data selection. I will mention a few below:

1 9 25 2K 15

0 0 4 341 0

View Details

Bert de Vries @bertdv0

a year ago

Some more refs: for implementation, rxinfer.com (toolbox), and arxiv.org/abs/2504.14898 (AIF planning as inference), and lazydynamics.com (company). Also check out noumenal.ai , stanhopeai.com, and verses.ai .

0 0 5 270 1

View Details

Bert de Vries @bertdv0

a year ago

(6) Finally, FEP is more than a pretty good idea as it can be derived from first principles by information theory, see e.g., blog at t.ly/Bl2DO plus refs. An AIF process avoids ad hoc design choices often found in man-made AI algorithms.