Google released EmbeddingGemma 2 to enable on-device search across text, code, images, video, and audio using a shared embedding space and modular encoders. The model reaches 740 million parameters and is tested at about 567MB of active RAM on a Pixel 11 Pro. Developers can deploy smaller embedding indexes by truncating vectors to 512/256/128 dimensions without retraining and add new modalities later without re-embedding existing data, with LiteRT/MediaPipe Tasks now and Android ML Kit integration coming soon.
Scrimshaw Jukebox used Claude Opus 5.5 to generate computer game music in a text-based format and created an artifact to play it aloud. The write-up says the music was aimed at matching the quality of the original Secret of Monkey Island. It changes by showing the output can be surprisingly good, prompting questions about whether music composition is a recent new capability of text models.
Falcon-Emirati-7B was released as a Falcon-H1-Arabic-based LLM specialized for understanding and generating Emirati Arabic rather than defaulting to Modern Standard Arabic. It scored 84.83% on the Alyah Emirati-dialect benchmark and, in open-ended judging, returned in Emirati dialect with a 0.52 partial-credit dialect-fidelity score versus 0.05 for the next best model. The result is that models using this approach better match Emirati tone and cultural nuance across Alyah categories, and they more reliably switch registers when asked in Emirati.
Every AI story that matters,
in your inbox by 8am.
TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the
day in two minutes. Follow companies and topics for alerts, or get the
briefing in Slack. Free, no spam, unsubscribe anytime.
Reading TLDRocket needs no cookies, and the readership counts we rely on come from
our own cookieless analytics. Google Analytics is the exception: it sets cookies and
reports to Google, so it stays switched off until you allow it. You can change your
mind any time from “Cookie settings” in the footer.
Strictly necessary
Session security and form protection (tldrocket-session,
XSRF-TOKEN, 2 hours). The site cannot work without them,
so they need no consent.
Always on
Google Analytics 4 (_ga,
_ga_<id>, up to 2 years). Measures which
stories and sections readers use. Google acts as a third-party processor and may
store the data outside the EU. No advertising, no profiling, no data sold.
