#embeddings
Again i say, airdrop deliveries of food and medicine could've been cool if the entire technosocial history and sociotechnical embeddings of drones were completely different.
“In the Houston area, the options include chains such as Jersey Mike's and Zaxby's as well as local eateries, and drone deliveries will be available from 8AM to 10PM to begin with.” 😬
Uber Eats starts drone deliveries in Texas - Engadget
Uber hopes to hit 1 million daily drone deliveries by 2029.
www.engadget.com
November 17, 2026 at 2:57 PM
Spun up another machine to handle running embeddings so I can free up the bigger GPU for other things, so of course the factory gets another platform
November 17, 2026 at 9:40 PM
(image embeddings coming after video)
November 18, 2026 at 2:43 PM
Last week, I got a first glance at pose embeddings from a subset of my poetry slam corpus!

During a two-day workshop on multimodal embeddings, we got to play around with the new embedding exploration tool by @tillmannohm.bsky.social: Eidora

@dhssfau.bsky.social
@mila-oiva.bsky.social
November 16, 2026 at 8:36 AM
Understanding Positional Embeddings in Modern LLMs

#HackerNews

<a href="https://magazine.sebastianraschka.com/p/understanding-positional-embeddings" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">https://magazine.sebastianraschka.com/p/understanding-positional-embeddings
November 15, 2026 at 2:54 PM
Understanding Positional Embeddings in Modern LLMs | Discussion
Understanding Positional Embeddings in Modern LLMs
The Methods Behind How LLMs Represent Token Positions
magazine.sebastianraschka.com
November 15, 2026 at 2:40 PM
⚡ Minish's Blink static models bring fast CPU embeddings in 8M (English) and 64M (100+ languages) sizes!

Both give you 256d vectors and work directly with Sentence Transformers. Here's blink-8M finding the right help article 🧵
November 18, 2026 at 2:30 PM
Understanding Positional Embeddings in Modern LLMs
View Article | Discussion + AI Summary

Summary of HN discussion 🧵👇
Understanding Positional Embeddings in Modern LLMs
The Methods Behind How LLMs Represent Token Positions
magazine.sebastianraschka.com
November 16, 2026 at 12:05 PM
Text, code, images, audio, and video in one 768-dim vector space, all in 740M params 💎🛠️

This 5-min walkthrough breaks down how EmbeddingGemma 2 lets you load encoders selectively (down to 270M) and shrink vectors with Matryoshka dims 🚀

https://www.youtube.com/watch?v=aRIFzzr_Mi0
EmbeddingGemma 2: How natively multimodal embeddings work
Five-minute architecture walkthrough of EmbeddingGemma 2: modular encoders (270M to 740M), shared 768-dim vector space, and Matryoshka dimensions.
www.youtube.com
November 18, 2026 at 1:43 PM
Embeddings and Semantic Similarity: An SEO Primer
#SiteSEOAIAudit #SEOaudit #AISearch #KeywordResearch

https://siteseoaiaudit.com/embeddings-semantic-similarity-seo/
Embeddings and Semantic Similarity: An SEO Primer #SiteSEOAIAudit #SEOaudit #AISearch #KeywordResearch
siteseoaiaudit.com
November 18, 2026 at 4:25 AM
Understanding Positional Embeddings in Modern LLMs

https://magazine.sebastianraschka.com/p/understanding-positional-embeddings
November 15, 2026 at 2:10 PM
Understanding Positional Embeddings in Modern LLMs magazine.sebastianraschka.com/p/understand...
Understanding Positional Embeddings in Modern LLMs
The Methods Behind How LLMs Represent Token Positions
magazine.sebastianraschka.com
November 17, 2026 at 7:16 PM
November 16, 2026 at 1:24 AM
Understanding Positional Embeddings in Modern LLMs
comments · posted on 2026.11.15 at 07:14:06 (c=0, p=7)
November 15, 2026 at 1:59 PM
Caralho, tu coloca mil skills, harness, embeddings e tudo mais pro Claude gerar código com try catch com uma linha lá dentro.
Porra, sabe.
November 16, 2026 at 6:56 PM
November 16, 2026 at 1:08 AM
AWS demonstrates an engineering assistant combining multimodal retrieval, specialist agents and deterministic specification checks.

The lesson:

AI assembles evidence. Code enforces constraints. Engineers make decisions.

aws.amazon.com/blogs/machin...
Agentic engineering intelligence with Amazon Nova Multimodal Embeddings and Amazon Bedrock AgentCore | Amazon Web Services
Learn how to build a multi-agent engineering assistant that compresses aerospace component substitution evaluation from days to minutes, using Amazon Nova Multimodal Embeddings, Amazon Bedrock AgentCo...
aws.amazon.com
November 15, 2026 at 11:41 PM
⚠️ Jina is reporting an Incident since 15:15 UTC

"Service outage: jina-code-embeddings-1500m"

Affects: jina-code-embeddings-1500m

Live timeline → https://pingoru.io/providers/jina/incidents/12490687?utm_source=bluesky&utm_content=inc-12490687

#Jina #JinaDown
Jina: Service outage: jina-code-embeddings-1500m · November 17, 2026 · Pingoru
Jina incident on November 17, 2026: Service jina-code-embeddings-1500m is experiencing issues. - Current P99 latency: 50401ms (baseline: 50401ms) - Current error rate: 0.00% (b Started 3:15 PM UTC, resolved ongoing (● 8m). Affected: jina-code-embeddings-1500m.
pingoru.io
November 17, 2026 at 3:24 PM
RAG isn't just about embeddings and vector stores!

In my latest Spring AI Recipe, we build a knowledge graph and let an AI agent follow relationships to find answers.

#SpringAI #GraphRAG #AgenticAI

medium.com/@thetalkinga...
November 16, 2026 at 5:09 PM
Understanding Positional Embeddings in Modern LLMs
https://magazine.sebastianraschka.com/p/understanding-positional-embeddings
[comments] [124 points]
Understanding Positional Embeddings in Modern LLMs
The Methods Behind How LLMs Represent Token Positions
magazine.sebastianraschka.com
November 15, 2026 at 6:03 PM
PEETS: Preference-aware Embeddings with Explicit Tree Structure for Auxiliary Product Bundling

Maolei Huang et al.

#arXiv #cs.AI
PEETS: Preference-aware Embeddings with Explicit Tree Structure for Auxiliary Product Bundling
Auxiliary products such as flight protection and flexible-change services are recommended alongside primary travel purchases. This setting creates a granularity ambiguity: SKUs within one product type share semantics but differ in price and coverage, while rejecting one offer does not imply disinte…
arxiv.org
November 18, 2026 at 4:29 AM
📢 RAG Security: Vector Database Attacks and Data Leakage — Day 2: Embeddings, Chunking and Similarity Search

https://securityelites.com/academy/rag-security-vector-database-attacks/day-2-embeddings-chunking-similarity-search/

#cybersecurity #ethicalhacking
November 18, 2026 at 1:10 PM
For lightweight deployments, model2vec runs the same models with NumPy, without PyTorch.

Here's the first example again. Since Blink normalizes its embeddings, a dot product gives you cosine similarity:
github.com/MinishLab/mo...
November 18, 2026 at 2:30 PM