Le projet THINK

TensorFlow* developers can benefit from Intel® Advanced Matrix Extension (AMX) with TensorFlow 2.12

Businesses pressured to adopt generative AI due to benefits, but hurdles exist, especially for enterprises. Prediction Guard, Intel® Liftoff member, highlights LLM model issues: unreliable, unstructured output hindering system development. Integrations raise legal, security concerns: output variability, compliance gaps, IP/PII leaks, « injection » vulnerabilities.

Intel had six papers accepted at the 40th International Conference on Machine Learning (ICML) 2023, happening now through July 29. Two papers were selected as spotlight oral papers at the conference: ProtST, which uses a ChatGPT-style design interface for protein design, and Settling the Reward Hypothesis, which explores the design of goals in reinforcement learning.

Intel and Mila collaborated on ProtST, a multi-modal protein language model that enables users to create protein sequences and structures based on natural language prompts. The ProtST research paper was presented in a prestigious spotlight talk at this year’s International Conference on Machine Learning (ICML 2023).

Introducing: Intel® QuickAssist Technology (Intel® QAT) an external sequence producer for zstd

Developer Spotlight: Akash Agrawal proposed an Object Detection Model for Autonomous Vehicles

Developer Spotlight: Rohit Sroch, in his blog proposed a Learning Enhancement and Assistance Platform

Intel Liftoff for Startups presents the AI Innovation Bridge: An onsite LLM and Gen AI hackathon during the Intel Innovation 2023 conference, fostering groundbreaking advancements. The event promises a transformative experience for AI startups. Set for September 19-20 in San Jose, the hackathon aims to unleash developers’ creative prowess with access to cutting-edge Intel resources.

Intel® oneAPI Threading Building Blocks: open-source library on GitHub supports parallelism on CPUs.

Using oneDNN for Improved Efficiency and Sustainability