Alessandro Betti retweeted
#LOT Spring school
@biggiobattista presenting the lecture "Machine Learning Security: Lessons Learned and Future Challenges"
@Cossu94 @v_lomonaco @StefanoMelacci @aleitteb
Alessandro Betti retweeted
๐ Learning Over Time (LOT) Spring School ๐
24-27 March 2025 | ๐ Siena, Italy
๐sites.google.com/unisi.it/loโฆ
Organizers: @StefanoMelacci ,@v_lomonaco @Cossu94 ,@aleitteb @TiezziMatteo
#ContinualLearning #CL #LifelongLearning #CollectionlessAI
Alessandro Betti retweeted
๐ Weโve launched a new site for Collectionless AI, sharing our take on next-gen AI/ML! ๐ Explore our papers, events, and more!
cai.diism.unisi.it/
#CollectionlessAI #ContinualLearning #LifelongLearning #MachineLearning #AI ๐
@StefanoMelacci @KleineBottleM @TiezziMatteo
Alessandro Betti retweeted
#CoLLas2024 folks ๐จ
Temperature rising for the coming sessions ๐ฅ๐จ
Alessandro Betti retweeted
Hurry up for @CoLLAs_Conf registration!
We strongly encourage #CoLLAs2024 participants to register before July 15th - which is only a few days away. Sign up today and start packing for Pisa! ๐
Alessandro Betti retweeted
๐ข "State-space Modeling in Long Sequence Processing: A Survey on Recurrence in the Transformer Era"
arxiv.org/abs/2406.09062
a joint work by @TiezziMatteo , M. Casoni, @aleitteb , M.Gori (@KleineBottleM) and @StefanoMelacci
Alessandro Betti retweeted
We also summarized the most common #benchmarks and settings for long sequence processing and #longrange dependencies arxiv.org/abs/2406.09062
Alessandro Betti retweeted
3โฃ RNNs evolutions, such as xLSTMs (@gklambauer), ODE-inspired RNNs (Neural Oscillators - @tk_rusch)
as well as novel paths for
4โฃ Learning in RNNs, such as modern evolutions of RTRL (@KhurramJaved_96) or bio-inspired methods (@BerenMillidge)
arxiv.org/abs/2406.09062
Alessandro Betti retweeted
2๏ธโฃ Deep State-space Models #SSMs such as #S4, #S5, #Mamba, #Griffin etc.
a line of work by @_albertgu @tri_dao @SamuelMLSmith, @orvieto_antonio and many others
arxiv.org/abs/2406.09062
Alessandro Betti retweeted
We described several trends:
1๏ธโฃ #Transformers embraced Recurrence exploiting a stateful recurrent representation:
*๏ธโฃ all the latest models RWKV4/5/6 by @BlinkDL_AI
*๏ธโฃ #RetNet by @donglixp et al.
are competitive with quadratic self-attention!
arxiv.org/abs/2406.09062
Alessandro Betti retweeted
Goal: efficiently ๐ฉ๐ซ๐จ๐๐๐ฌ๐ฌ ๐ฅ๐จ๐ง๐ ๐ฌ๐๐ช๐ฎ๐๐ง๐๐๐ฌ + ๐ฉ๐ซ๐๐ฌ๐๐ซ๐ฏ๐ ๐ฅ๐จ๐ง๐ ๐ญ๐๐ซ๐ฆ ๐๐๐ฉ๐๐ง๐๐๐ง๐๐ข๐๐ฌ
TL;DR #Transformers + #RNNs, #DeepStatespace models โก๏ธ progressive confluence of previously disjoint architectures
arxiv.org/abs/2406.09062
Alessandro Betti retweeted
๐งInterested in the details of the latest SOTA architectures (#RWKV , #RetNet, #Mamba, #Griffin) for long sequences?
๐ข "State-space Modeling in Long Sequence Processing: A Survey on Recurrence in the Transformer Era" arxiv.org/abs/2406.09062
Alessandro Betti retweeted
๐ขJay McClelland, co-author of the seminal 'Parallel Distributed Processing' book ๐, is set to be our keynote speaker at #CoLLas2024! ๐
Alessandro Betti retweeted
โผ๏ธShort survey paperโผ๏ธ
๐ข"On the Resurgence of Recurrent Models for Long Sequences: Survey and Research Opportunities in the Transformer Era"
by @TiezziMatteo M.Casoni @aleitteb @_masaccio_ @KleineBottleM @StefanoMelacci
#Transformers #SSMs
arxiv.org/abs/2402.08132
Alessandro Betti retweeted
๐จ1 month left until the #CoLLAs2024 paper submission deadline!๐จ
Donโt miss your opportunity to contribute to the field of lifelong learning. Finalize your research and submit by 15 Feb 2024!
๐ Prepare your submission: lifelong-ml.cc/Conferences/2โฆ
๐๏ธ Abstract deadline: 09 Feb 2024
Alessandro Betti retweeted
It is sad to lose the DeepMind office in Edmonton to the Tech layoffs and looming recession. But AI is not going away, and I am more focused than ever on the Alberta Plan for AI research.
arxiv.org/abs/2208.11173
Alessandro Betti retweeted
๐ฃ Last week I presented "Foveated Neural Computation" @ECMLPKDD
โก๏ธ Paper: 2022.ecmlpkdd.org/wp-contentโฆ
โก๏ธ Code: github.com/sailab-code/foveaโฆ
A thread๐๐งต
In some visual tasks, such as image classification, salient information is distributed in regions of limited size ๐ง
1/n
Alessandro Betti retweeted
(8/n) @TiezziMatteo, @MarulloSimone, Lapo Faggi, @enricomeloni_ai, @aleitteb and @StefanoMelacci investigated the effectiveness of gravitational models in an open-set class-incremental setting, by imposing stochastic coherence along the visual scanpath. arxiv.org/abs/2204.12193
Alessandro Betti retweeted
Mi raccomando, durante gli incontri parentali delle feste, oltre a stare attenti, non appena lo zio dice โsiero speri..โ siate saggi ๐
Alessandro Betti retweeted
Italy's new profession of "queuer" was set to reduce queues. It may increase them econ.st/1pETIr5
