One of nine recognized this year: https://
aclweb.org/portal/content
/acl-fellows-2024
…
RESEARCH
-
Gautam Kamath Recognized as ACL Fellow 2024
By
–
-
Jimmy Lin Named 2024 ACL Fellow for QA and IR
By
–
Congratulations to my awesome @UWCheritonCS colleague Jimmy Lin (
@lintool
) on being named a 2024 ACL Fellow! "For significant contributions to question answering and information retrieval." Right on the heels of being named a 2022 ACM Fellow… he's on a roll! -
Domain-Specific Evaluation Sets for LLM-as-a-Judge Pipeline
By
–
Devs, let's chat at @NeurIPSConf about a previous paper: Constructing Domain-Specific Evaluation Sets for LLM-as-a-judge. The idea is to introduce a data pipeline to generate domain-specific evals for LLM-as-a-Judge, for uses cases. Paper
-
Data Explosion Accelerates: 197-Fold Growth by 2028
By
–
197-Fold #Data Explosion, #AI Eat Away! 2010: 2 ZB
2015: 15.5 ZB
2020: 64 ZB
2023: 123 ZB
2024: 149 ZB
2026: 221 ZB projected
2028: 394 ZB projected #IoT #Streaming #BigData @StatistaCharts -
Gemini vs Perplexity: Speed and Success
By
–
Comparing Gemini Deep Research with Perplexity Pro Reasoning. Neither of the solutions manager to solve the task but:
— 🚨 AI News | TestingCatalog (@testingcatalog) 11 décembre 2024
– Perplexity is much much faster
– Deep Reasoning was able to complete it after 5 attempts https://t.co/pnT8Ltfqti pic.twitter.com/zIMl67k49eComparing Gemini Deep Research with Perplexity Pro Reasoning. Neither of the solutions manager to solve the task but: – Perplexity is much much faster – Deep Reasoning was able to complete it after 5 attempts
-
Responsible AI in the Intelligence Age: Ethical Considerations
By
–
Thank you so much my @ThuliumCo team for sharing my article and amplifying this important conversation! The Intelligence Age brings immense potential, but it also calls for thoughtful reflection on how we harness AI responsibly. I'm grateful to be part of a team that values both
-
SambaLingo: Teaching LLMs New Languages at NeurIPS
By
–
Chat with us at @NeurIPSConf about our SambaLingo paper. #NeurIPS2024 It's all about teaching LLMs new languages & diving deeper into how we train our SambaLingo series of multilingual models. Paper Link https://
bit.ly/3YROYAl -
Publishing research in Statistics over CS for community engagement
By
–
Thank you! We probably could have published it faster in a CS conference. But I really think it's of significant interest to the Statistics community, and believe there should be more engagement between the CS and Stats privacy communities, so we put our money where our mouth is.
-
Cerebras Trains Trillion-Parameter Models on Single CS-3
By
–
Learn more here: https://
cerebras.ai/press-release/
cerebras-demonstrates-trillion-parameter-model-training-on-a-single-cs-3-system
… https://
cerebras.ai/blog/introduci
ng-gigagpt-gpt-3-sized-models-in-565-lines-of-code
… -
Linear Scaling AI Model Training Across Multiple Systems
By
–
After the single box ran, we scaled it to 2 and 16 systems with linear scaling. There were no model & code changes required.
Megatron
DeepSpeed
Sharding The model lives on a single block of memory. No need to break it up. The whole thing trains like one giant GPU.