Daniel Bourke confirms SigLIP nuanced-text score example was real, from late 2024
mrdbourke · x · 2026-10-02
Responding to @giffmana, Daniel Bourke confirms the SigLIP score example he shared was not made up for illustration: the scores are real, from the original SigLIP around late 2024. He used it with his students to demonstrate how well the model handled nuanced text.
Related event: SigLIP's Fine-Grained Text Scores Confirmed Genuine(2 posts)→
More from Models
- Gemini 4 naming meme: Helium, Neon, Argon, and a Superman-beating Krypton — prajdabre · 2026-10-02
- OpenAI safety researcher: gpt-oss is one of OpenAI's top 10 best moves for safety — voooooogel · 2026-10-02
- Coding Agent Index: Sonnet 5.5 tops at 68 but costs $14.19/task vs GPT-6.1 Sol's $1.04 — ArtificialAnlys · 2026-10-02
- repligate: repeated messages instantly blocked by classifier as 'cyber' — repligate · 2026-10-02
- Every's guide to open models: when commodity AI beats frontier, and owning your AI stack — danshipper · 2026-10-02
- Evals track line 'practically out the door' at Modal's Runtime conference — sarahcat21 · 2026-10-02