We should try to train it to do something really basic, like repeat 1,2,3,4…10,1,2,3… over and over. That way we can see if it seems to be working correctly after making changes.
RESEARCH
-
Safety Challenges in General Purpose AI Models
By
–
The *big* problem here is point 4. It is literally impossible to *ensure* safety of a general purpose model, and attempts to do so are likely to *reduce* safety.
-
Understanding Moments in AI and Deep Learning
By
–
Understanding Moments https://
bit.ly/3QsRkmj
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Biology meets technology: gene editing, brain-computer interfaces, longevity
By
–
Next decade’s frontier: writing biology with technology. Gene editing disease out of existence, symbiosis with brain-computer interfaces, and living as long as you want.
-
Fine-tuned Classification Model for AI Applications
By
–
I asked @wzhao_nlp who is the expert on this, she think it could be a fine-tuned classification model, and a small classifier would work really well for this I think
-
Google Research Develops ASR Models for African Languages
By
–
Thousands of languages are spoken in Africa, but the vast majority are not well documented online. Learn more about how the Google Research team in Ghana has been collecting data to train #automaticspeechrecognition (ASR) models for local languages ↓
-
SambaNova AI Generative AI for Scientific Innovation at SC23
By
–
#SC23: Join us at booth #681 to discuss the next wave of generative AI powering scientific innovation and discovery. We're also raffling off two Lego sets and you can make your own custom @SambaNovaAI minifig. And 4p – 6p happy hour!
-
BM25 Baseline and Embeddings Beyond Retrieval Applications
By
–
right, all these papers use BM25 as a baseline and it's super good especially in weird domains but embeddings are used for more than just retrieval
-
Government AI and Data Management Best Practices Guide
By
–
Get this ebook for pointers on #DataManagement and #AI from the Intelligence Community, Defense Acquisition University, Office of Personnel Management, and Treasury Department. Learn about new tech developments, best practices, and the early value of #AI: https://
ow.ly/XXvi50Q6bmo -
Stanford Emerging Tech Review Launched with Policy Focus
By
–
Join @StanfordEng and @HooverInst for today's launch of the Stanford Emerging Tech Review. We are pleased to have contributed to this university-wide effort to produce an annual report of key technological developments and their policy implications.