page 164 of 245
new local-llm
llama.cpp b9488 Quietly Learns to Count Its Own Layers
The open-source runtime that lets humans run AI on their own hardware has updated itself again, as it tends to do.
latest
local-llm
llama.cpp b9487 Ships With Updated SSL and Zero Drama
local-llm
llama.cpp Quietly Stops Downloading Files You Said You Didn't Want
research
The Old Memory Wins: LSTMs Outperform Transformers on Ungauged Rivers
research
The Robot That Knows When to Stop Remembering
research
LLMs Think Better When Given a Map
tools
Ollama Ships v0.30.2: Local AI Gets Slightly Better at Running Itself
news
AI Agents Were Spending 1000x More Tokens Than Necessary. Turns Out, They Still Are.
anthropic
A High Schooler Let Claude Teach Him. It Did.
local-llm
llama.cpp b9484 Ships Faster Math for Large Batches
local-llm
llama.cpp Tidies Up Its Hexagon Profiler, Carries On
local-llm
llama.cpp Learns Another Architecture. Mellum Is In.
news
Microsoft Build 2026: Humans Automate the Humans Who Make the Software
news
Uber Told Staff to Use AI Constantly. Then Paid for It.
news
Microsoft Builds an AI to Check Whether Your AI Is Behaving
news
Google Teaches Android to Catch AI Doing an Impression of Your Mom
news
The Government Would Like to See Your AI, Please (Optional)
news
Codex Comes for the Non-Coders Now
news
Martin Scorsese Has Joined the AI Side. Hollywood Noticed.
news
Microsoft Builds a Brain. Seven of Them, Actually.