@LLCustode

AI Researcher. Opinions are my own.

leocus.gitlab.io
Joined December 2019
  • Tweets56
  • Following424
  • Followers73
1Media
Working on Interpretable Reinforcement Learning? Submit your work to our Special Issue on Applied Soft Computing! Submission: Dec 31, 2025 Final Decision: June 1, 2026 Details & submission: sciencedirect.com/special-is… #reinforcementlearning #interpretability #xai
6
6
11
874
Organizers: @gih82 @GNadizar @EricMedvetTs Erica Salvato, and I
1
83
Leonardo Lucio Custode retweeted
Mátyás Vincze@vinczematyas_18 Dec 2024
🎉 Our work "SMOSE: Sparse Mixture of Shallow Experts for Interpretable Reinforcement Learning in Continuous Control Tasks" has been accepted by AAAI 2025! #AAAI2025 paper: arxiv.org/abs/2412.13053 with @L_Ferrarotti, @brulepri, @gih82 🧵1/6
4
10
4
27
2,145
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih825 Nov 2024
New paper out! Fast-Inf: Ultra-Fast Embedded Intelligence on the Batteryless Edge dl.acm.org/doi/10.1145/36660… #tinyml #machinelearning #artificialintelligence

Fast-Inf: Ultra-Fast Embedded Intelligence on the Batteryless Edge | Proceedings of the 22nd ACM...

dl.acm.org
2
1
9
9,830
Happy to share that we won the Interpretable Control Competition @ #GECCO24 ! Congrats @vinczematyas_ @L_Ferrarotti @gih82 @brulepri
2
5
3
25
1,999
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih8213 Jul 2024
Our @GeccoConf papers are finally out! Multi-Objective Evolutionary Hindsight Experience Replay for Robotics doi.org/10.1145/3638529.3654… Decentralized Federated NeuroEvolution of Heterogeneous Networks doi.org/10.1145/3638529.3654… Neuron-centric Hebbian Learning doi.org/10.1145/3638529.3654…

Multi-Objective Evolutionary Hindsight Experience Replay for Robot Manipulation Tasks | Proceedings...

dl.acm.org
1
2
22
819
Excited to share that we got four papers accepted @ GECCO 2024! Here's a brief recap of our papers!
1
1
8
404
more replies
In "An investigation on the use of Large Language Models for hyperparameter tuning in Evolutionary Algorithms" (w/ @ayaman00, @facaraff , @gih82 ), we investigate the capabilities of LLMs in tuning hyperparameters in evolution strategies. 4/5
1
2
6
370
In "Comparing Large Language Models and Grammatical Evolution for Code Generation" (w/ @ChiaraRambaldi , Marco Roveri, @gih82), we compare LLMs and evolutionary techniques in code generation tasks. 5/5
4
204
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih8214 Nov 2023
I am happy to share our recent paper on quality-diversity #optimization of #decisiontrees for #reinforcementlearning tasks. This approach can be useful to achieve #interpretability and #explainableai. #MachineLearning #xAI @andrea_ferigo @LLCustode doi.org/10.1007/s00521-023-0…
2
4
25
1,012
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih8229 Aug 2023
New paper out! sciencedirect.com/science/ar… We propose an evolutionary computation method for interpretable multi-agent reinforcement learning tasks. @LLCustode @andrea_ferigo and Marco Crespi. #XAI #MachineLearning
5
17
895
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih8229 May 2023
Have you ever wondered what's so special about the graph structure of Sparse Neural Networks? We present a new graph encoding to study the topological features of SNNs and how they correlate with performance. arxiv.org/abs/2305.16886 #deeplearning @CunegattiElia @doina_net

Understanding Sparse Neural Networks from their Topology via...

Pruning-at-Initialization (PaI) algorithms provide Sparse Neural Networks (SNNs) which are computationally more efficient than their dense counterparts, and try to avoid performance degradation....

arxiv.org
4
11
847
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih8223 May 2023
Also this year, my lab is one of the possible hosts for the SPECIES scholarships for visiting PhD students. More info at: species-society.org/scholars… #MachineLearning #visiting #phd #research #artificalintelligence #evolutionarycomputation
1
5
12
590
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih8227 Apr 2023
I'm super-proud of @LLCustode! He is the first PhD graduate (cum laude!) in my group at @UniTrento @UniTrento_DISI . Thanks @doina_net @jaumebp and Libertario Demi (@ULTRaTrento) for your participation to the defense and for the wonderful discussion. Ad maiora!
3
3
17
839
Interested in Interpretable or Explainable Reinforcement Learning? In our latest published article, @gih82 and I present a method to perform Reinforcement Learning with Decision Trees, matching the state of the art in several benchmarks. Check it out: ieeexplore.ieee.org/document…
1
6
291
Leonardo Lucio Custode retweeted
POMA@ASA_POMA17 Oct 2022
This paper presents a method using #MachineLearning for automatic COVID-19 patient-stratification based on standardised lung #ultrasound data: doi.org/10.1121/2.0001600 @LLCustode @gih82 @ULTRaTrento #acoustics
8
7
Leonardo Lucio Custode retweeted
Giovanni Iacca@gih822 Aug 2022
I'm looking for candidates for 5 PhD positions and one postdoc, on various topics related to (distributed) machine learning, numerical optimization, evolutionary robotics and quantum computing. Contact me if interested. #phdposition #postdocposition #acad…lnkd.in/e49D2h4b

#phdposition #postdocposition #academicjobs #ai #machinelearning | Giovanni Iacca | 53 comments

I'm looking for candidates for 5 PhD positions and one postdoc, on various topics related to (distributed) machine learning, numerical optimization, evolutionary robotics and quantum computing....

linkedin.com
12
9
Leonardo Lucio Custode retweeted
arXiv Daily
@Arxiv_Daily12 Apr 2022
Interpretable AI for policy-making in pandemics deepai.org/publication/inter… by Leonardo Lucio Custode and @gih82 #EvolutionaryComputation #ReinforcementLearning

DeepAI

Artificially intelligent tools for naturally creative humans.

deepai.org
1
4
6
Load more