70
2
Qwen3-VL 2B on Raspberry Pi with llama.cpp (eheidi.dev)
1
Apple and Foxconn in China (damagemag.com)
3
Toward provably private insights into AI use (research.google)
1
Xyne: Open-source LLM-driven search engine for Google Workspace (github.com/xynehq)
1
People Use ChatGPT (nber.org)
1
Playground Elements: Serverless interactive editable coding env for the web (github.com/google)
1
Salesforce near membrane: Sandbox JavaScript object graphs (salesforce.com)
1
Meta Clip 2: Worldwide (arxiv.org)
2
RAG Anti-Patterns (jxnl.co)
1
Code Signing for web apps using Signed HTTP Exchange (2023) (rot256.dev)
2
Havard's private companies are collectively worth $393B (phoenixfundresearch.substack.com)
2
Small Language Models: Techniques, Enhancements, Applications, Trustworthiness (arxiv.org)
57
Guide to Fine-Tuning LLMs (arxiv.org)
1
Updated NotebookLM allows for more control (webwizwork.com)
1
Technical Goals (2016) (openai.com)
2
Vpqy: ML framework for queryable video analytics (github.com/vqpy)
1
Altitude: Triage violent, extermist, and terrorist content (github.com/jigsaw-code)
1
Meta's Agentic System: Multistep reasoning,websearch,code interpreter for Llama3 (github.com/meta-llama)
1
Coppie: Rewrite content using 232 copywriting formulas with ChatGPT (github.com/wynterjones)
11
Mutlimodal neural networks converge to a shared statistical model of reality (arxiv.org)
4
Tell HN: Manara (YC W21) seeking companies to commit hiring engs from Palestine (manara.tech)
74
Machine Unlearning in 2024 (stanford.edu)
57
GPT-4 can exploit vulnerabilities by reading CVEs (theregister.com)
3
Running fine-tuned LoRA models on Workers AI (cloudflare.com)
1
Sqlelf: Explore ELF Objects through SQL (github.com/fzakaria)
1
WikiLLM: Inject facts into LLMs by editing parameters (github.com/laramohan)
2
Shimshady: Linux Secure Boot CVE (eclypsium.com)
1
Jina.ai: Self-host Multimodal models (github.com/jina-ai)
1
Conversational search engine with Bing and Lepton.ai (github.com/leptonai)
0
A survey of resource-efficient LLMs and multimodal models (arxiv.org)
143
My AI costs went from $100 to less than $1/day: Fine-tuning Mixtral with GPT4 (twitter.com/wenquai)
2
Augmenting LLMs with Knowledge: A survey on hallucination prevention (arxiv.org)
160
Vanna.ai: Chat with your SQL database (github.com/vanna-ai)
144
Mixtral 8x7B: A sparse Mixture of Experts language model (arxiv.org)
1
LLMs cannot self-correct reasoning yet (arxiv.org)
1
3 Static Queue (FIFO) cache eviction algorithm
5
The race to spend IRA's $100B in climate grants (heatmap.news)
1
Cohere Chat API with Retrieval-Augmented Generation in Public Beta (cohere.com)
1
Social Media x Political Dysfunction: Interference, manipulation, misinformation (docs.google.com)
2
LeanStore: OLTP storage engine optimized for many-core CPUs and NVMe SSDs (github.com/leanstore)
5
The many vulnerabilities in SOHO and industrial wireless routers (talosintelligence.com)
1
Secure Programmable Router (github.com/spr-networks)
6
The Wizard of Oz Approach (2016) (ycombinator.com)
1
Search-agumented instruction learning: LLaMA1-7B fine-tuned with web search (github.com/luohongyin)
1
Optimising compute, video compression, & code generation with generalised AI (deepmind.com)
1
WebGLM: Web-Enhanced Q&A with LLMs (github.com/thudm)
1
Aidan Gomez: List of best arguments against AI doomerism (twitter.com/aidangomezzz)
1
Letting employees set their own pay and its results (frappe.io)
1
The first group affiliated with YC Research is OpenAI (2015) (ycombinator.com)
3
Patrick Collison Interviews Sam Altman on the Future of AI (youtube.com)
53
Michael Schumacher's family take legal action over AI-generated interview (skysports.com)
2
Pieces of Advice (calvinrosser.com)
1
Rocks, pebble, sand: A work sprint prioritisation framework (asmartbear.com)
26
On “I don't trust microcode” (2021) (georgi.family)
23
Composefs: Content-Addressable Overlay Filesystem for Linux (github.com/containers)
2
Salus: A micro hypervisor for RISC-V systems (github.com/rivosinc)
1
Helping medical doctors and patients in the Foundation Model AI era (erictopol.substack.com)
1
On the opportunities and risks of foundation AI models (arxiv.org)
1
Eric Barone & his one-person game studio has made $300M revenue in 6yrs (twitter.com/volodarik)
145
The Need to Read (paulgraham.com)
32
[flagged] A person I consider trustworthy is convinced crypto will face a systemic risk (twitter.com/paulg)
4
The progress principle: Power of small wins (2011) (hbr.org)
2
Immich: Self-hosted photos and videos backup for Android and iOS (github.com/immich-app)
3