Fair but it’s still actively learning to make those errors (just so it can recover from them later), which imo is still a bit weird. In an ideal world you wouldn’t, possibly process supervision is one way to get there even within RL framework. Def agree on “recovery learning”.
MACHINE LEARNING
-
TMLR Papers Eligible for NeurIPS ICML ICLR Presentation
By
–
We at TMLR are proud to announce that selected papers will now be eligible for an opportunity to present at the joint NeurIPS/ICML/ICLR Journal-to-Conference (J2C) Track:
-
Data Mixing Balance: Vision-Language Skills and Regional Imbalance
By
–
That's a great question and will take the opportunity to dash off a bit on the data mixing. Mixing data is a tricky balance it turns out. There were two main factors at play: – we wanted to keep general vision-language skills.
– and had unbalanced regions and languages: think -
Training multilingual cultural model improves multimodal benchmark performance
By
–
Thank you so much and appreciate you taking time to check the paper. We trained a model(
https://
huggingface.co/neulab/Cultura
lPangea-7B
…) on the subset of the dataset and were able to improve performance on many multilingual/cultural multimodal benchmarks. Yeahhh, would be great too if you could also -
Scaling 2.8M Images and 3M Entities with Flexible Data Curation Pipeline
By
–
We have around 2.8M unique images and 3M unique entities and the data curation pipeline allow scaling that up(more images per entity, adding new countries/languages, questions per entity, etc…). Very glad to hear people were asking about this 🙂
-
Major VQA Dataset Release: 30M Samples Across 42 Countries
By
–
Congrats! Big props unifying vision language datasets and making it easy to access them. We also recently released a massive VQA dataset of 30M samples(22M filtered), spanning 42 countries and 39 languages. The dataset contains factual visual questions and answers about world
-
Physics-Informed AI Improves Molecular Structure Predictions
By
–
Physics-informed learning is essential for AI+Science. Check out our latest paper in @PNASNews special edition on AI+Chemistry on incorporating simple rules of physics that two atoms in a molecule cannot get too close to one another (due to repulsive forces) vastly improves
-
Complexity vs Feature Engineering Trade-offs in Machine Learning
By
–
The less complexity the more capacity to focus on learning other things. But the more handcrafted features the less opportunity to learn better representations. It all kind of depends on how big the model is, how much data you have, and what your compute resources are
-
New Tokenization Approach Challenges BPE-Style Methods
By
–
TLDR: surprised that it works, skeptical whether this will replace BPE-style tokenizers, delighted about this refreshing approach