Introducing An Evolved Universal Transformer Memory https://
sakana.ai/namm Neural Attention Memory Models (NAMMs) are a new kind of neural memory system for Transformers that not only boost their performance and efficiency but are also transferable to other foundation models,
LLMS
-
Neural Attention Memory Models Boost Transformer Performance
By
–
-
QwenCoder2.5 32B Now Available on SambaNova Cloud
By
–
QwenCoder2.5 32B now on Sambanova Cloud! #Devs using @continuedev can now experience lightning-fast code suggestions powered by @Alibaba_Qwen Coder 2.5 32B. QwenCoder2.5 32B in the coding #interface Supercharge your workflow now Learn more https://
bit.ly/3D217LI -
NeurIPS 2026: AI and Large Models Meet Protein Research
By
–
catch @thisismadani at neurips lots of ai + large models + proteins
-
Command Models Improve Tool Generation with Structured Outputs
By
–
Tool generations by Command models just got better! We’ve updated our Structured Outputs so that our models now strictly adhere to your tool descriptions, ensuring all required parameters are included with accurate data types.
-
Strict Tools Feature Added to API V2 Chat Calls
By
–
To try it out, just add strict_tools=True to your usual co . chat() tool calls in API V2. Learn more:
-
HourVideo: MLLM Benchmark Dataset for Long-Form Video Understanding
By
–
Please come to see this work called HourVideo, a benchmarking dataset to evaluate the capabilities of MLLMs’ understanding and reasoning of long form videos. https://t.co/EH7hwJsDwH
— Fei-Fei Li (@drfeifei) 9 décembre 2024Please come to see this work called HourVideo, a benchmarking dataset to evaluate the capabilities of MLLMs’ understanding and reasoning of long form videos.
-
Transformers Inference Optimization Toolset for AI
By
–
Transformers Inference Optimization Toolset https://
bit.ly/41gChC4
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
LLMs Benchmarking for Embodied Decision Making Work
By
–
Please come to see this work benchmarking LLMs for embodied decision making! https://t.co/8ogXa9pPEv
— Fei-Fei Li (@drfeifei) 9 décembre 2024Please come to see this work benchmarking LLMs for embodied decision making!
-
SubgoalXL Advances AI-Powered Theorem Proving Frontier
By
–
Our research with @HKUniversity
, the SubgoalXL is advancing the frontier of #AI-powered theorem proving. Explore the framework & results below GitHub:
https://
bit.ly/3ZnnFhI
Datasets & Models:
https://
bit.ly/3ZKF3gD
Paper:
https://
bit.ly/3ZHYCXS #NeurIPS2024 -
Alibaba Qwen 2.5-32B-Coder: Fast Open-Source Model Release
By
–
📣 @Alibaba_Qwen 2.5-32B-Coder now available!
— SambaNova (@SambaNovaAI) 9 décembre 2024
✅ Access the best open-source models
✅ 5X faster than any #GPU provider
We've partnered with @continuedev to make it easier for #devs to integrate these models & experience fast #inference response times.
See how: 👇@Alibaba_Qwen 2.5-32B-Coder now available! Access the best open-source models 5X faster than any #GPU provider We've partnered with @continuedev to make it easier for #devs to integrate these models & experience fast #inference response times. See how: