Skip to content
Tech News
clear
Topics: Today This Week This Month This Year
1.
Just brute force your embeddings (news.ycombinator.com)
2.
Miniaturizing and modifying natural proteins with Raygun (feeds.nature.com)
3.
AI developer runs 28.9-million-parameter model on $10 ESP32-S3 microcontroller — uses Google's Per-Layer Embeddings technique, stores table on 16MB Flash memory (tomshardware.com)
4.
Truth is not a direction: a Tarski attack on LLM probes (news.ycombinator.com)
5.
Running a 28.9M parameter LLM on an $8 microcontroller (news.ycombinator.com)
6.
Universal cell embedding provides a foundation model for cell biology (feeds.nature.com)
7.
Dispersion loss counteracts embedding condensation in small language models (news.ycombinator.com)
8.
Show HN: CLI tool for detecting non-exact code duplication with embedding models (news.ycombinator.com)
9.
Gemini API File Search is now multimodal (news.ycombinator.com)
10.
RAG precision tuning can quietly cut retrieval accuracy by 40%, putting agentic pipelines at risk (venturebeat.com)
11.
Show HN: Atomic – Local-first, AI-augmented personal knowledge base (news.ycombinator.com)
12.
Servo is now available on crates.io (news.ycombinator.com)
13.
Solving Semantle with the Wrong Embeddings (news.ycombinator.com)
14.
Google's Gemini Embedding 2 arrives with native multimodal support to cut costs and speed up your enterprise data stack (venturebeat.com)
15.
Querying 3B Vectors (news.ycombinator.com)
16.
Connes Embedding Problem (news.ycombinator.com)
17.
Why MongoDB thinks better retrieval — not bigger models — is the key to trustworthy enterprise AI (venturebeat.com)
18.
ChromaDB Explorer (news.ycombinator.com)
19.
Using Vectorize to build an unreasonably good search engine in 160 lines of code (news.ycombinator.com)
20.
Prompt caching for cheaper LLM tokens (news.ycombinator.com)
21.
Prompt caching: 10x cheaper LLM tokens, but how? (news.ycombinator.com)
22.
Contextualization Machines (news.ycombinator.com)
23.
28M Hacker News comments as vector embedding search dataset (news.ycombinator.com)
24.
Why Google’s File Search could displace DIY RAG stacks in the enterprise (venturebeat.com)
25.
Ilo – a Forth system running on UEFI (news.ycombinator.com)
26.
Meta Superintelligence Labs' first paper is about RAG (news.ycombinator.com)
27.
Meta Superintelligence's surprising first paper (news.ycombinator.com)
28.
Language models pack billions of concepts into 12k dimensions (news.ycombinator.com)
29.
Language Models Pack Billions of Concepts into 12k Dimensions (news.ycombinator.com)
30.
Language Models Pack Billions of Concepts into 12,000 Dimensions (news.ycombinator.com)
Today's top topics: power nasa apple instruments officials science devices google messages burned
View all today's topics →