DEV Community

#llm

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
Agent Memory: Everything It Remembers Has the Same Authority, and That Is the Bug

Comments suggest policies need human promotion

Agent Memory: Everything It Remembers Has the Same Authority, and That Is the Bug

2
Comments 5
5 min read
Deploying a QAT Checkpoint Your Serving Stack Can't Load: Gemma 4 E2B in Pure JAX on One TPU

Deploying a QAT Checkpoint Your Serving Stack Can't Load: Gemma 4 E2B in Pure JAX on One TPU

2
Comments
10 min read
Qwen3.8-27B: A Deep Dive Into Qwen's Newest Vision-Language Powerhouse

Qwen3.8-27B: A Deep Dive Into Qwen's Newest Vision-Language Powerhouse

8
Comments 2
6 min read
Five agent engineering problems, with the numbers behind them

Five agent engineering problems, with the numbers behind them

Comments
6 min read
The smoke test said yes

The smoke test said yes

Comments
12 min read
Why AI Should Return Interfaces, Not Walls of Text

Why AI Should Return Interfaces, Not Walls of Text

Comments
3 min read
My agent had two rules and obeyed neither

My agent had two rules and obeyed neither

Comments
9 min read
Our Documentation Was Lying. The Model Believed It.

Our Documentation Was Lying. The Model Believed It.

Comments
8 min read
Hyperia 0.17.5 is Released

Hyperia 0.17.5 is Released

Comments
1 min read
The Geometry Measures, the AI Teaches: Building an Art Studio Tutor with ADK, Vertex AI & OpenCV

The Geometry Measures, the AI Teaches: Building an Art Studio Tutor with ADK, Vertex AI & OpenCV

Comments
3 min read
Yorishiro: macOS Terminal Hosting AI Agents

Yorishiro: macOS Terminal Hosting AI Agents

Comments
2 min read
Elevated Errors on Claude Opus 5 Impacting Automation Workflows

Elevated Errors on Claude Opus 5 Impacting Automation Workflows

Comments
2 min read
From Chatbot to Production AI System: What Changes When AI Meets Real Software

From Chatbot to Production AI System: What Changes When AI Meets Real Software

Comments
5 min read
Debian Starts Voting on AI/LLM Contributions for Future Development

Debian Starts Voting on AI/LLM Contributions for Future Development

Comments
2 min read
I labelled 300 broken LLM tool calls - passing them straight to the schema fails 290 of 300

I labelled 300 broken LLM tool calls - passing them straight to the schema fails 290 of 300

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