Cohere Labs Community

community

https://cohere.com/research

Cohere_Labs

Cohere-Labs-Community

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

tellarin authored a paper 1 day ago

LectūraAgents: A Multi-Agent Framework for Adaptive Personalized AI-Assisted Learning and Embodied Teaching

tellarin submitted a paper 2 days ago

LectūraAgents: A Multi-Agent Framework for Adaptive Personalized AI-Assisted Learning and Embodied Teaching

jjzha authored a paper 21 days ago

CommonLID: Re-evaluating State-of-the-Art Language Identification Performance on Web Data

View all activity

Reubencf

posted an update about 7 hours ago

Post

Tiny Maple is now live on iOS and Android.

The app introduces the @Coherelabs Tiny Aya series of multilingual AI models to mobile devices. This release is significant as it enhances access to multilingual AI from anywhere, particularly for users who prefer offline capabilities.

I invite you to try it and share your feedback.

App Store: https://apps.apple.com/bn/app/tiny-maple/id6774123088

Google Play: https://play.google.com/store/apps/details?id=com.reubencf.tinyaya

mmhamdy

posted an update 3 days ago

Post

2516

What if you could train a model on just 10 images instead of 60,000 and still get close to the same performance?

Traditional machine learning requires thousands, even millions, of data points to achieve high accuracy. But what if we could "distill" the entire dataset into just a few synthetic samples?

This is what Dataset Distillation offers. Unlike traditional knowledge distillation, we keep the model fixed and distill the knowledge contained in a massive training set into a tiny set of synthetic distilled images.

The goal is to train a model on this ultra-small set and achieve performance that almost matches what the same model would get when trained on the massive original dataset.

For example, training on only 10 distilled MNIST images (this is equivalent to a single image per class) yields 94% accuracy, compared to 99% when training on the full 60,000 images.

Interestingly, these distilled images look significantly different (as you can see in the image below) from natural images because they are optimized for model training rather than for matching the correct data distribution.

But that's not all.

Most importantly, this same method opens the door to a potent form of data poisoning. Because distilled images are specifically optimized for rapid learning, an attacker can create a tiny set of adversarial distilled images to cause a well-trained model to forget or misclassify a specific category.

What I find fascinating about dataset distillation is this: it mimics human-like learning by letting a model grasp a concept from a single example, but it does so using alien synthetic images that mean absolutely nothing to a human eye!

What about you? What are your thoughts on it?

2 replies

Reubencf

posted an update 8 days ago

Post

2034

Millions speak Konkani. The internet barely knows it.

Today's major LLMs struggle with regional languages. They can't read, write or even recognize Konkani. So I built one that can.

Here is a working demo of the Konkani LLM I've been training. 👇

https://youtu.be/8K04ylbXh6k

mmhamdy

posted an update 9 days ago

Post

4951

It was supposed to be a failed experiment. Instead, it led to the discovery of one of the most intriguing phenomena in neural networks, simply because a researcher forgot to turn it off and left it running....for a week!

In 2022, researchers at OpenAI were studying how neural networks generalize from their training data. For this task, they were training small transformer models to perform modular arithmetic.

The thing is, neural networks are weird. When a model has an abundance of parameters (like neural nets), it can easily overfit. It essentially memorizes its training data, scoring a perfect 100% accuracy when tested on it, but remains completely clueless when faced with any new instances not present in the training set (close to 0 accuracy). It is like memorizing 1 + 2 = 3 without understanding the concept of addition, so if 2 + 3 wasn't in the training set, the model fails miserably!

Usually, when a model overfits like this, people just cut their losses, turn off the experiment, and move on with their lives.

But sometimes they forget. And that is exactly what happened to our researchers at OpenAI. A week later, they checked back in, and a miracle had happened!

They discovered Grokking (And no, this has nothing to do with xAI's Grok , the term was originally coined by sci-fi author Robert Heinlein to mean understanding something so deeply that it becomes part of you). Grokking is when a neural network suddenly and abruptly learns to generalize long after it has overfitted. Just take a look at the graph in the image below!

Spooky, right! I told you neural nets are weird!

4 replies

mmhamdy

posted an update 12 days ago

Post

1010

Human brains don't recreate every pixel to understand the world!

Most current models in genomics, proteomics, and single-cell transcriptomics rely on generative objectives like masked language modeling or next token prediction. While effective, these architectures waste significant capacity reconstructing raw, noisy sequence details that may not carry functional biological meaning.

But a promising, more efficient alternative is emerging: Joint-Embedding Predictive Architecture (JEPA)

Originally introduced by Yann LeCun for computer vision, JEPA is a non-generative, self-supervised learning (SSL) framework. Instead of predicting raw inputs, it operates as a world model that predicts abstract semantic embeddings in latent space.

Recently, the JEPA framework (and its more efficient LeJEPA variant) has been adapted into the biological sciences to develop performing foundation models and to improve on already existing ones.

It's interesting how each adaptation modified and tailored JEPA to suit its specific biological domain, whether by experimenting with different backbones or complementing the objective with other loss terms.

For example, JEPA-DNA and ProteinJEPA used JEPA as a continual pre-training framework to enhance existing foundation models without training from scratch, while Cell-JEPA and JEPA-DNA employed a hybrid objective that combines the JEPA loss with a traditional language modeling loss.

The article below provides an overview of these implementations, along with others that came out this year. As always, your thoughts and feedback are welcome and highly appreciated!

Link to the article is in the first comment 👇

3 replies

bryanlimy

authored a paper 13 days ago

Wearable data from subjects playing Super Mario, sitting university exams, or performing physical exercise help detect acute mood episodes via self-supervised learning

Paper • 2311.04215 • Published Nov 7, 2023

mmhamdy

posted an update 20 days ago

Post

136

Things rarely go as we expect!

In 2017, Google released the Transformer architecture. While it was clear the model was promising, absolutely no one (including its authors) anticipated the pervasive global revolution it would create!

The authors actually viewed the Transformer as just a stepping stone for a much more ambitious project: The MultiModel.

Their ultimate goal was to build a single deep learning architecture capable of jointly learning massive, diverse tasks across entirely different domains (in 2017). A One Model To Learn Them All.

In fact, the MultiModel paper was published in the exact same month as Attention Is All You Need!

But history had other plans. The building block eclipsed the grand design!

So, have you heard about the MultiModel before? 😀