⚖ optimizing for consensus - hugonoss · Scour

Flash Attention 2 in CuteDSL: A Naive Kernel, Three Optimizations, and Where I Got Stuck 🚀Performance

kyrieblunders.bearblog.dev·6d·Hacker News

Incompressible Knowledge Probes: Estimating Black-Box LLM Parameter Counts via Factual Capacity 🦙Ollama

arxiv.org·2d·Hacker News

Cryptography challenges @KalmarCTF 2026 💡New and interesting problems

blog.zksecurity.xyz·3d·Hacker News

Topics in High-Performance Messaging 🚀Performance

ultramessaging.github.io·5d·Lobsters

Studying the emergence of leaders in moving crowds of pedestrians 📈analyse historical interactions to predict future preferences

On the Limits of Self-Improving in Large Language Models: The Singularity Is Not Near Without Symbolic Model Synthesis 🦙Ollama

arxiv.org·2d·Lobsters

TurboQuant: A First-Principles Walkthrough 🧮Vector Databases

taoofmac.com·3d

Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses 🦙Ollama

arxiv.org·2d·Hacker News

Show HN: WaveletLM – wavelet-based, attention-free model with O(n log n) scaling 🧮Vector Databases

github.com·4d·Hacker News

Guess-Verify-Refine: Data-Aware Top-K for Sparse-Attention Decoding on Blackwell via Temporal Correlation 🧮Vector Databases

arxiv.org·4d·Hacker News

GPT-5.5 is a biased evaluator: authorship and order effects 🚀Performance

blog.valmont.dev·6d·Hacker News

2026-04-30 🔮near or far future

parentheticalrecluse.com·5h

No more posts from hugonoss's subscribed feeds.

Scour all 24433 feeds Learn more about Feeds

Log in to enable infinite scrolling