Md Rifat Arefin

PhD Candidate @mila_quebec @UMontreal | Research Scientist Intern @Meta | Prev. @Samsung @Amazon @ServiceNowRSRCH @RecursionPharma @withupstride

[Md Rifat Arefin @rarefin15]

@fereshte_khani Want to chat as well! Can't dm you!

[Md Rifat Arefin @rarefin15]

@JohnCLangford Would love to chat, can not dm you!

[Patrick Huber @huberpa91]

1/n 🚀 Introducing MobileLLM-Pro: Our latest 1 billion parameter foundational language model designed to deliver high-quality, efficient on-device inference across a wide range of general language modeling tasks – A thread 🧵

🤗Model Checkpoints: https://t.co/t6ocm3wGSB

Work done with my amazing collaborators Ernie Chang, Wei Wen, Igor Fedorov, Chinnadhurai Sankar, Tarek Elgamal, Hanxian Huang, Naveen Suda, Vish Vogeti, Yanghan Wang, Alex Gladkov, Kai Sheng Tai, Zechun Liu, Abdelrahman Elogeel, Tarek Hefny, Yangyang Shi, Vikas Chandra, Ahmed Aly, Anuj Kumar, Raghuraman Krishnamoorthi, and Adithya Sagar

[Irina Rish @irinarish]

I am looking for a postdoc to lead projects related to this collaboration, on scaling laws, emergence and interpretability in pre- and post-training & inference/reasoning, in multimodal foundation models (language, time series, tabular data etc). HPC experience is a plus.

[Md Rifat Arefin @rarefin15]

Come see our poster and talk if you are at ICML!

[Ravid Shwartz Ziv @ziv_ravid]

If you're at ICML, come tomorrow (Tuesday) to Oscar's talk, where he will present our paper "Layer by layer: Uncovering hidden representations in language models" at 10am (West Ballroom D) and for the poster session at 11am (East Exhibition Hall A-B #E-2607).

[Md Rifat Arefin @rarefin15]

@johnsdvorak @ziv_ravid @nicogontier @ylecun @irinarish @chrisjpal Hi, thanks for your interest! You can check Appendix A in the OpenReview version (https://t.co/4pvjJj0g4m). We'll be updating the arXiv version soon!

[Ravid Shwartz Ziv @ziv_ravid]

Interested in representation in intermediate layers of LLMs? What is a good representation, and how can it be measured? Visit Oscar's poster on "Does Representation Matter? Exploring Intermediate Layers in Large Language Models"

[Md Rifat Arefin @rarefin15]

@PetarV_93 @GoogleDeepMind Congratulations Petar!

[Md Rifat Arefin @rarefin15]

Join the world's largest and diverse community of academic researchers in deep learning. Apply by December 1!

[Mila - Institut québécois d'IA @Mila_Quebec]

Mila's annual supervision request process opens on October 15 to receive MSc and PhD applications for Fall 2025 admission! Join our community! More information here https://t.co/r01eLcXtZw

[Md Rifat Arefin @rarefin15]

Excited to share our work "Unsupervised Concept Discovery Mitigates Spurious Correlations" at #ICML2024
Come see our poster today, July 25th at 1h30-3pm.

Paper: https://t.co/cBHRuyxKnv
Poster: https://t.co/HPGIkC5kql
Code: https://t.co/tu4d4aT5a5

@Mila_Quebec

[Irina Rish @irinarish]

🚨 Open-source AI community - stop building everything from scratch, let's build on each other's (and your own) work over time - continually, as we should! Don't waste previoius compute and human effort!

See https://t.co/jugB0DW4KZ - simple & useful tips on how to just keep training foundaiton models on new data without forgetting or performance degradations w.r.t. usual large-scale training!

Kudos to our amazing team, @kgupta24 @benjamintherien @ai_phd @MatsRichter @QuentinAnthon15 @TLesort @ebelilov !

[Irina Rish @irinarish]

Join our Scaling, Alignment & Open-Source AI workshop Fri, Dec 15 in New Orleans or remotely https://t.co/3lsCLcVBqV!
Hear Nick Bostrom, Yoshua Bengio (TBA) @ylecun @tegmark @percyliang @juliabossmann @JJitsev @ethanCaballero @irinarish and others discuss the Open Source and the Future of AI, and engage in discussions on new scaling laws, "emergent" behaviors, continued pretraining, multimodal alignment, compression & fast inference, time-series foundation models and open generalist agents, and recent supercomputing efforts (Trillion Parameter Consortium) and more! Join us for the Open Source AI party sponsored by @cerebras, @togethercompute and @UMontreal's CERC-AAI Lab led by @irinarish!

[Md Rifat Arefin @rarefin15]

@FrancescoLocat8 @ISTAustria Congrats Francesco!

[Irina Rish @irinarish]

Great job, @togethercompute and INCITE (@UMontreal @AiEleuther @laion_ai ) teams and all the collaborators! It's awesome to see the RPI 7B (RedPajama-INCITE) model outperform ALL the best open models of same size - Llama-7B, Falcon-7B, and MPT-7B on HELM benchmark!

[Timothée Lesort @TLesort]

We are glad to announce that our paper: "Challenging Common Assumptions about Catastrophic Forgetting" co-authored with @ostap__alex, @__z__9, @rarefin15, @prlz77 , @lcharlin , @irinarish has been accepted to @Collas2023.
(arXiv https://t.co/HHbczhnyCU)
[A thread (1/N)]