9/ Making LLMs Better at Dense Retrieval – proposes LLaRA which adapts an LLM for dense retrieval; LLaMa-2-7B was improved on benchmarks like MSMARCO and BEIR.
MACHINE LEARNING
-
Survey of Reasoning with Foundation Models: Latest Advancements
By
–
8/ A Survey of Reasoning with Foundation Models – provides a comprehensive survey of seminal foundational models for reasoning, highlighting the latest advancements in various reasoning tasks, methods, benchmarks, and potential future directions.
-
MLP Layers as Lookup Tables for Factual Recall in LLMs
By
–
5/ Fact Recalling in LLMs – studies how MLP layers implement a lookup table for factual recall; suggests that early MLP layers act as a lookup table and recommends thinking about the recall of factual knowledge in the model as multi-token embeddings.
-
Survey of 300+ Papers: Generative AI from Gemini to Q-Star
By
–
2/ From Gemini to Q-Star – surveys 300+ papers and summarizes research developments to look at in the space of Generative AI; it covers computational challenges, scalability, real-world implications, and more.
-
Innovative Multimodal Models Redefine AI Capabilities
By
–
7. Innovative multimodal Models • LLAMA 2, SAM, LLaVA, and more redefine image segmentation, language vision, and advanced Al capabilities.
-
PaLM-E: AI Robotics Breakthrough in Multimodal Interpretation
By
–
5. AI in Robotics with PaLM-E • PaLM-E blends image, text, and robotics, showcasing Al's capability to interpret and interact with the real world.
-
Fast Inference of Mixture-of-Experts Language Models with Offloading
By
–
Fast Inference of Mixture-of-Experts Language Models with Offloading Artyom Eliseev, Denis Mazur : https://
arxiv.org/abs/2312.17238 #ArtificialIntelligence #DeepLearning #MachineLearning -
2024 AI Predictions Thread: Share Your Forecasts
By
–
Okay, 2024 prediction thread for AI. Leave your predictions below! I'll check this tweet next year to see who had the best predictions. 🙂
-
Named Tensors Utility Without Complete Coverage Requirements
By
–
yes its here to stay, as it doesn't need complete coverage to be useful (unlike named tensors)
-
functorch.dim Upstreamed to PyTorch Official Repository
By
–
functorch.dim and its docs.
used to be here but it's upstreamed now to functorch.dim