[go: up one dir, main page]

Forem

# llm

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
OpenAI Structured Outputs vs Zod: which to use for LLM response validation in 2026

OpenAI Structured Outputs vs Zod: which to use for LLM response validation in 2026

Comments
3 min read
AI Agents That Learn on the Job: Why On-the-Fly Evolution Changes Everything About Agent Architecture
Cover image for AI Agents That Learn on the Job: Why On-the-Fly Evolution Changes Everything About Agent Architecture

AI Agents That Learn on the Job: Why On-the-Fly Evolution Changes Everything About Agent Architecture

Comments
3 min read
Getting Qwen and Gemma to play Zork (and why they get stuck in the maze)

Getting Qwen and Gemma to play Zork (and why they get stuck in the maze)

Comments
7 min read
A Vector Database Is Not a RAG Pipeline -And Confusing the Two Will Cost You

A Vector Database Is Not a RAG Pipeline -And Confusing the Two Will Cost You

Comments
7 min read
Your AI Agent Gets Dumber the More You Teach It. Skill Graphs Are the Fix.

Your AI Agent Gets Dumber the More You Teach It. Skill Graphs Are the Fix.

Comments
3 min read
Ollama, LM Studio, and GPT4All Are All Just llama.cpp — Here's Why Performance Still Differs

Ollama, LM Studio, and GPT4All Are All Just llama.cpp — Here's Why Performance Still Differs

Comments
6 min read
Prompt Injection Doesn't Come from Your Users
Cover image for Prompt Injection Doesn't Come from Your Users

Prompt Injection Doesn't Come from Your Users

Comments
10 min read
Why Prompt-Based Agents Don’t Scale (and What We’re Trying Instead)

Why Prompt-Based Agents Don’t Scale (and What We’re Trying Instead)

Comments 2
2 min read
Per-customer cost attribution without a proxy

Per-customer cost attribution without a proxy

Comments
3 min read
Anthropic Just Did Something Unprecedented: They Kept a Model Because It Was Too Good at Hacking

Anthropic Just Did Something Unprecedented: They Kept a Model Because It Was Too Good at Hacking

Comments
3 min read
GLM-5.1: The 754B Open Model That Writes Animated SVG

GLM-5.1: The 754B Open Model That Writes Animated SVG

Comments
1 min read
The Chinese Open-Source Model That Draws Pelicans Better Than GPT-4o

The Chinese Open-Source Model That Draws Pelicans Better Than GPT-4o

Comments
2 min read
I Benchmarked the Viral "Caveman" Prompt to Save LLM Tokens. Then My 6-Line Version Beat It.
Cover image for I Benchmarked the Viral "Caveman" Prompt to Save LLM Tokens. Then My 6-Line Version Beat It.

I Benchmarked the Viral "Caveman" Prompt to Save LLM Tokens. Then My 6-Line Version Beat It.

Comments
5 min read
"Parenting" a Language Model

"Parenting" a Language Model

Comments
4 min read
LLM-as-Judge: using Claude to review a Gemini agent

LLM-as-Judge: using Claude to review a Gemini agent

Comments
7 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.