Top ML Papers of the Week (Oct 30 – Nov 5): – LLMs for Chip Design
– Battle of the Backbones
– Next Generation AlphaFold
– Symmetry in Machine Learning
– Enhancing LLMs by Emotion Stimuli
– Efficient Context Window Extension of LLMs
…
GENERATIVE AI
-
Top ML Papers: LLMs, AlphaFold, and Advanced Techniques
By
–
-
Fair Use and AI Training: Legal Arguments and Copyright Implications
By
–
Notice how this particular legal argument goes against what Anthropic is arguing. Fair Use implies it does fall under Copyright… What's missing from this analysis is that the determination of Fair Use depends on the application itself, whether it's harmful to rightsholders.
-
Emergence of Specialized Domain-Specific LLMs in Finance and Law
By
–
In some very specific cases, we will also see some very specialized LLMs emerge – e.g. financeLLM or LegalLLM. These domain-specific LLMs that may require a lot of custom training, RLHF, and fine-tuning. /16
-
Smaller LLMs with RAG as cost-effective alternative to GPT-4
By
–
Smaller and more efficient LLMs that have reasonably good reasoning capabilities can be fine-turned or complemented with RAG and incorporated into the workflow to automate these use cases. Using GPT-4 or some other very large LLM will become cost-prohibitive in these cases. /14
-
Price-Competitive API Calls for General-Purpose LLM Use Cases
By
–
General-purpose use-case embedded in a product or service For these, a vanilla API call to SOTA LLM is sufficient. Price will be the key consideration here and as long as the large LLM providers have very competitive prices, making simple calls to their APIs will work. /12
-
Enterprise AI and LLM APIs: Two Classes of Business Use Cases
By
–
Enterprise AI and LLM APIs The other big category of LLM use cases is businesses using these LLMs in their core products, services, and business processes. There are 2 classes of use cases in this space. /11
-
Future AI Services Market: Limited Competition with Subscription Models
By
–
It is extremely unlikely that we will have more than 2-3 of these services. These services, like ChatGPT, will have a free and paid subscription tier. Paying subscribers will enjoy premium features like personalized responses and access to multi-modal features etc. /10
-
Google, Meta, and OpenAI dominate consumer LLM market
By
–
All this means that we will end up with Google, Meta, and potentially OpenAI being the key players in the consumer LLM (e.g., ChatGPT, Bard, etc.) world. /9
-
General-Purpose vs. Specialized Tasks: When Fine-Tuning and RAG Matter
By
–
It's important to draw the distinction between general-purpose tasks like Python code generation vs. specialized tasks like having knowledge of the Abacus platform. The former typically DOES NOT require fine-tuning or RAG, while the latter requires some custom work. /8
-
GPT-4 Outperforms Specialized LLMs Across Tasks
By
–
This is unlikely to be the case for general-purpose tasks. Large SOTA LLMs outperform specialized LLMs in most tasks. Again GPT-4 outperforms specialized LLMs on pretty much everything from code generation to writing and reasoning tasks. /7