All tags
Person: "karina-nguyen"
Contextual Document Embeddings: `cde-small-v1`
llama-3 cde-small-v1 gemini-1.5-flash-8b chatgpt meta-ai-fair openai google-deepmind weights-biases togethercompute contextual-embeddings contextual-batching video-generation synthetic-data model-efficiency training-techniques rag algorithmic-efficiency jack-morris sasha-rush tim-brooks demis-hassabis karina-nguyen
Meta announced a new text-to-video model, Movie Gen, claiming superior adaptation of Llama 3 to video generation compared to OpenAI's Sora Diffusion Transformers, though no release is available yet. Researchers Jack Morris and Sasha Rush introduced the cde-small-v1 model with a novel contextual batching training technique and contextual embeddings, achieving strong performance with only 143M parameters. OpenAI launched Canvas, a collaborative interface for ChatGPT with synthetic data training. Google DeepMind welcomed Tim Brooks to work on video generation and world simulators. Google released Gemini 1.5 Flash-8B, improving cost and rate limits with algorithmic efficiency.
Canvas: OpenAI's answer to Claude Artifacts
gpt-4o claude-artifacts openai cursor_ai daily inline-suggestions collaborative-editing code-editing model-training model-integration feature-detection accuracy-evaluation voice-ai hackathon open-source-libraries marijn-haverbeke karina-nguyen vicente-silveira swyx
OpenAI released Canvas, an enhanced writing and coding tool based on GPT-4o, featuring inline suggestions, seamless editing, and a collaborative environment. Early feedback compares it to Cursor and Claude Artifacts, noting strengths and some execution issues. OpenAI also sponsors Marijn Haverbeke, creator of ProseMirror and CodeMirror, which are used in Canvas. The integration involved training a detector to trigger Canvas appropriately, achieving 83% accuracy in correct triggers. Unlike Claude Artifacts, Canvas currently lacks Mermaid Diagrams and HTML preview support. Additionally, Daily is sponsoring a $20,000 voice AI hackathon in San Francisco, highlighting voice AI as a key emerging skill.
LMSys advances Llama 3 eval analysis
llama-3-70b llama-3 claude-3-sonnet alphafold-3 lmsys openai google-deepmind isomorphic-labs benchmarking model-behavior prompt-complexity model-specification molecular-structure-prediction performance-analysis leaderboards demis-hassabis sam-altman miranda-murati karina-nguyen joanne-jang john-schulman
LMSys is enhancing LLM evaluation by categorizing performance across 8 query subcategories and 7 prompt complexity levels, revealing uneven strengths in models like Llama-3-70b. DeepMind released AlphaFold 3, advancing molecular structure prediction with holistic modeling of protein-DNA-RNA complexes, impacting biology and genetics research. OpenAI introduced the Model Spec, a public standard to clarify model behavior and tuning, inviting community feedback and aiming for models to learn directly from it. Llama 3 has reached top leaderboard positions on LMSys, nearly matching Claude-3-sonnet in performance, with notable variations on complex prompts. The analysis highlights the evolving landscape of model benchmarking and behavior shaping.