Walid Saba
@sabawalid
PhD Comp Science (AI/NLU); 25 years experience (AT&T Bell Labs, IBM, and AIR); 40+ publications including an award winning paper
ID: 172737374
https://medium.com/ontologik 30-07-2010 13:37:46
887 Tweet
1,1K Takipçi
1,1K Takip Edilen
What is a Large Language Model? Is it accurate to say it "models" language? In a seminar last week, Walid Saba detailed how #LLMs model "statistical regularities" in language, and detailed "bottom-up reverse-engineering of language at scale." Read more: bit.ly/3vLhRUv
"When Colorless Green DNNs Sleep Furiously in an Unexplainable Fantasy," by Walid Saba, asks for the addition of sanity to discussion of LLMs, and for serious consideration of alternatives that could produce explainable, reliable, and scalable AI. bit.ly/4bJMwkl
As hype around #AI, #deeplearning and #LLMs continues to grow, are the systems' limitations being ignored? In a blog post for Association for Computing Machinery, Walid Saba, a senior research scientist at our institute, explored misguided claims being made about the #tech: bit.ly/3V8vYNi
Gary Marcus "How to spend billions building an unpredictable, hallucinating, unexplainable machine"
"No Generalization without Understanding and Explanation," by Walid Saba (Institute for Experiential AI at Northeastern Northeastern U.), says #LLMs will never be able to perform high-level reasoning like what's needed in deep language understanding, #ProblemSolving, and #planning. bit.ly/3BcQogK
