Good and Evil and Other Stories by Samanta Schweblin review – grasping the essence of horror
theguardian.com·10h
🏛Digital humanities
Using the Golden Ratio to Construct Poems
poetrywithmathematics.blogspot.com·1d·
Discuss: poetrywithmathematics.blogspot.com
📜TEI Encoding
weeknotes, 34/25
meiadeleite.com·2d
⭐Awesome lists
Andrés Vázquez: ‘90% of technological problems can be solved with tools that have been around for decades’
blog.okfn.org·45m
⚖️Lossy Compression Ethics
Can large language models figure out the real world?
news.mit.edu·20h
🧮Kolmogorov Bounds
The Relativity of Wrong by Isaac Asimov (1988)
hermiene.net·4h·
Discuss: Hacker News
🏛️Philosophy
You need a kitchen slide rule
entropicthoughts.com·58m·
Discuss: Hacker News
🧮Algorithmic Cooking
Building things of worth in a shallow world
ryanglover.net·1d·
Discuss: Hacker News
⬜Minimalism
Wubular: Rubular Reimagined in Ruby+WASM
rubyelders.com·2h·
Discuss: Hacker News
🌀Brotli Internals
Build and deploy small AI models from the browser
blog.minibase.ai·1d·
Discuss: Hacker News
🌀Brotli Internals
It's okay to solve a problem twice
til.andrew-quinn.me·3d·
Discuss: Hacker News
⚡Proof Automation
Reformulating web documents into synthetic data addresses the growing limits of AI training data
the-decoder.com·2d
📄Document AI
Constrained Prompt Enhancement for Improving Zero-Shot Generalization of Vision-Language Models
arxiv.org·12h
🧠Neural Codecs
Limits of message passing for node classification: How class-bottlenecks restrict signal-to-noise ratio
arxiv.org·12h
⧗Information Bottleneck
So You Want to Build an Open Source Alternative to ChatGPT for Teams
dev.to·4h·
Discuss: DEV
🔌Interface Evolution
Semantic Embedding in RAG: why close vectors still miss meaning and how to fix it
dev.to·11h·
Discuss: DEV
🧮Vector Embeddings
Benchmarking GPT-5 vs Claude 4 Sonnet on 200 Requests
dev.to·1h·
Discuss: DEV
⚙️Compression Benchmarking
Inaccuracy of Ensemble-Based Covariance Propagation, Beyond Sampling Error
arxiv.org·1d
📐Linear Algebra
DAIQ: Auditing Demographic Attribute Inference from Question in LLMs
arxiv.org·1d
💻Local LLMs
CE-RS-SBCIT A Novel Channel Enhanced Hybrid CNN Transformer with Residual, Spatial, and Boundary-Aware Learning for Brain Tumor MRI Analysis
arxiv.org·12h
🧠Machine Learning