DEV Community

#llm

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
Gemma 4 E2B on a Single TPU v6e Chip: A Serving Deep Dive

Gemma 4 E2B on a Single TPU v6e Chip: A Serving Deep Dive

10
Comments 1
8 min read
Your company already runs the hardest parts of an AI agent platform

Your company already runs the hardest parts of an AI agent platform

3
Comments
9 min read
Why You Should Never Let an LLM Decide Your AI Agent's Permissions

Why You Should Never Let an LLM Decide Your AI Agent's Permissions

Comments
4 min read
Why Unknown Model Pricing Should Fail Closed in AI-Agent Runtimes

Why Unknown Model Pricing Should Fail Closed in AI-Agent Runtimes

Comments
3 min read
Andrej Karpathy's "Deep Dive into LLMs like ChatGPT"

Andrej Karpathy's "Deep Dive into LLMs like ChatGPT"

5
Comments
4 min read
Agent Leaderboards Mislead Under Distribution Shift (IBM): Predictive Validity

Agent Leaderboards Mislead Under Distribution Shift (IBM): Predictive Validity

Comments
7 min read
codex fixing codex: a consensus loop that argues, judges, and merges its own PRs

codex fixing codex: a consensus loop that argues, judges, and merges its own PRs

Comments
5 min read
Catch LLM Schema Drift Before It Breaks Production

Catch LLM Schema Drift Before It Breaks Production

Comments
4 min read
SuperCompress is now on PyPI! pip install supercompress in 1 line

SuperCompress is now on PyPI! pip install supercompress in 1 line

Comments 1
1 min read
How to Prevent Prompt Injection in LangChain Python Apps

How to Prevent Prompt Injection in LangChain Python Apps

1
Comments
12 min read
SEO Isn't Dead — It Got a New Boss, and Its Name Is the Answer Box

SEO Isn't Dead — It Got a New Boss, and Its Name Is the Answer Box

Comments
4 min read
We fixed privacy-preserving local llm inference for developer tooling ? without a single API call.

We fixed privacy-preserving local llm inference for developer tooling ? without a single API call.

Comments
4 min read
Fail Closed, Not Open: Designing an AI Gateway for Regulated Enterprises

Fail Closed, Not Open: Designing an AI Gateway for Regulated Enterprises

Comments 1
7 min read
claude-docker; a sandbox for running Claude Code unattended

claude-docker; a sandbox for running Claude Code unattended

2
Comments 2
2 min read
Sakana AI's Fugu Explained: How the Multi-Agent Model Orchestrates Frontier LLMs

Sakana AI's Fugu Explained: How the Multi-Agent Model Orchestrates Frontier LLMs

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