🧠 Machine Learning - hop1.ng.1357 · Scour

You can put a Softmax in front of CrossEntropyLoss. PyTorch won’t stop you. Here are 16 other architecture bugs it won’t catch. ⚡Edge AI

gaox.substack.com·3d·Substack

Reconciling Contradictory Views on the Effectiveness of SFT in LLMs: An Interaction Perspective ✨LLMs

Andrej Karpathy Joined Anthropic. What It Says About Where AI Is Heading. 🛡️Anthropic PBC

firethering.com·19h·Hacker News

Andrej Karpathy joins Anthropic to accelerate Claude pre-training 🛡️Anthropic PBC

techthreedots.com·15h·Hacker News

Fine-Tuning NVIDIA Cosmos Predict 2.5 with LoRA/DoRA for Robot Video Generation ⚙️MLOps

huggingface.co·2d

The Ultimate LLM Fine-Tuning Guide ✨LLMs

promptinjection.net·3d·Hacker News

InferenceBench: A Benchmark for Open-Ended Inference Optimization by AI Agents 📱Edge AI Optimization

inferencebench.ai·4h·Hacker News

Show HN: Dataset for AI training and fine tuning ✨Gemini

neurvance.com·2d·Hacker News

dcostenco/prism-coder: The Mind Palace for AI Agents - HIPAA-hardened Cognitive Architecture with on-device LLM (prism-coder:7b), Hebbian learning, ACT-R spreading activation, adversarial evaluation, persistent memory, multi-agent Hivemind and visual dashboard. Zero API keys required. 🤖AI

github.com·11h·Hacker News

PyTorch, rewritten from scratch in pure Rust 🦋Tauri

github.com·6d·Hacker News

From Llama to Cria: Scaling Down Neural Networks via Neuron-Level Spectral Structural Importance Evaluation 🔢BitNet

Slax Reader CLI: a read-later library your AI agents can use 👨‍💻AI Coding

slax.com·21h·Hacker News

Embedding models are coordinate systems. What silently breaks in production RAG 📚RAG

internals.laxmena.com·1d·Hacker News

2D map of 26,741M/CV papers from CVPR, NeurIPS, ICML, ICLR (2024–2025) ⭐Content Scoring

matejgazda.com·6d·Hacker News

Running PyTorch Models on Apple Silicon GPUs with the ExecuTorch MLX Delegate ⚙️MLOps

pytorch.org·2d·Hacker News

Benchmarking Subquadratic's latest model and SSA Kernel 📱Edge AI Optimization

appen.com·6d·Hacker News

ArtifactLinker: Linking Scientific Artifacts for Automatic State-of-the-Art Discovery 🧠Obsidian

chiennv2000/orthrus: Fast, lossless LLM inference via dual-view diffusion decoding. 🤖LLM

github.com·5d·Hacker News

Google Scholar names its most influential papers for 2025 ⚡Edge AI

nature.com·1d·Hacker News

AI Slopification and Writing 🎭Claude

ordinaryintelligence.substack.com·6d·Substack

Log in to enable infinite scrolling