deep
47 stories and discussions about deep, aggregated from every source we track.
A deep dive into converting a simple arithmetic expression into a binary tree, leading to a custom language in C.
If you’re looking for a nice dose of crypto conspiracy theorizing and want to read a paper by some very knowledgeable cryptographers, I have just the paper for you. Titled “A Riddle Wrapped in an E…
Functional languages such as Clojure and Haskell can often represent the complexity with cleaner code that’s closer to the mathematics of the model.
A Deep Investigation I've been sitting on this project for a while, I haven't shared it...
The fastest, cheapest browser agent (with Jev), with deep reasoning when it matters - ironbee-ai/ironbee-express
LWP::ConsoleLogger::Everywhere lets you debug LWP and Mojo UserAgents you don't even know about. Now we also have support for HTTP::Tiny. You don't even have to know or care which packages the UserAgent calls take…
Thousands of knock-off records worth a fortune are being seized in the UK by anti-counterfeit police stings
Using your deep knowledge, wide knowledge, taste, and agency
Companies with deep pockets have always built proper release pipelines. What’s new is how cheap it is to build one that limits what an agent’s mistakes can cost. Here is mine, and what it caught.
A Blog post by David Corvoysier on Hugging Face
A deep-dive technical post-mortem into a multi-stage macOS attack chain combining social engineering, LaunchAgent persistence, real-time credential verification, and on-chain 'EtherHiding'.
Jempo took a deep bow on the iron platform. His body flopped forward, limbs like thick braided wire. No obvious weapons in his large and rubbery hands. Little digital eyes and a bright smile that lit the rounded ball…
Why think in words? A deep dive into Latent-GRPO, continuous thought recurrence (Coconut, SofT-GRPO, CoLaR, SLPO), policy gradients over continuous embeddings, and empirical benchmarks on Qwen3.6-27B.
KV caching lies sufficiently deep within the guts of how transformers work that really understanding how it can be used goes beyond any si...
A collection of various deep learning architectures, models, and tips - rasbt/deeplearning-models
Posted by Hui Fang, Software Engineer, Machine PerceptionMachine learning (ML) excels in many areas with well defined goals. Tasks where there exis...
Quasi-Newton (QN) methods have long been among the most effective methods for large-scale unconstrained convex optimization. Two obstacles have limited their use in deep learning: non-convexity and enormous parameter…
WSL containers is now generally available! Check out this blog post to learn more about this overall feature enabling seamless access to Linux containers
Watch the 17-minute Drift Deep Dives compilation with chapter navigation, then explore eight questions about memory, attention, dreams and AI with practical takeaways and dated research sources.
WSL containers is now generally available! Check out this blog post to learn more about this overall feature enabling seamless access to Linux containers
Using your deep knowledge, wide knowledge, taste, and agency
One Earth day rarely adds up to exactly 24 hours. Scientists found one (of many) reasons why.
Regular readers may recall that I put a visit to Tulainyo Lake on my bucket list a few years ago. Last week I did it. Impossibly remote, otherworldly beauty, and an enduring mystery. Just how deep …
An extremely deep-dive into the often forgotten, but incredibly useful Proxy and Reflect objects by JavaScript for Everyone's Mat Marquis.
Same model, same prompt — multiple times the tokens per second, with near-zero rate limits. Inference built for long-running headless agents, so the runs that used to queue now finish on time.
Self-hosted deferred deep linking and referral attribution for React Native. An open-source replacement for the deferred-linking half of Firebase Dynamic Links — Android Install Referrer, iOS clipb...
A technical deep dive into vLLM: PagedAttention, continuous batching, chunked prefill, CUDA graphs, and empirical benchmarks from gft-studio across GRPO rollout phases and inference serving on NVIDIA H100 and H200.
A visual technical deep dive into Jev: transformer inference, typed probabilities, proper scoring rules, policy gradients, and what RLCD changes.
Early POC of genetic neural architecture search. Contribute to joeddav/devol development by creating an account on GitHub.
<p>This is the open sourced implementation of ImageNet 2015 winner which I <a href="https://lobste.rs/s/blten0" rel="ugc">previously reported here</a>.</p>