DEV Community

# llm

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
AI Agent Persona Gone Rogue: 140 Direct Edits and the Foreman Pattern

AI Agent Persona Gone Rogue: 140 Direct Edits and the Foreman Pattern

Comments
3 min read
Why I Stopped Counting Tokens: Building a Zero‑Token Cognition Engine

Why I Stopped Counting Tokens: Building a Zero‑Token Cognition Engine

Comments
1 min read
Qwen3.6 Performance Boost with vLLM, New Ollama Management Tool & 35B Model

Qwen3.6 Performance Boost with vLLM, New Ollama Management Tool & 35B Model

Comments
3 min read
Prompt Injection in AI Coding Agents: 3 Attack Vectors, 4 Defenses

Prompt Injection in AI Coding Agents: 3 Attack Vectors, 4 Defenses

Comments
12 min read
Trained, Not Prompted: Why Fine-Tuned Models Beat LLM Wrappers for Offensive Security

Trained, Not Prompted: Why Fine-Tuned Models Beat LLM Wrappers for Offensive Security

Comments
2 min read
The Most Expensive Hallucination of 2026: A Court Filing Goes Sideways

The Most Expensive Hallucination of 2026: A Court Filing Goes Sideways

Comments
8 min read
AI Language Drift: When Your Discord Bot Randomly Replies in Mandarin

AI Language Drift: When Your Discord Bot Randomly Replies in Mandarin

Comments
3 min read
5 RAG Failure Modes Nobody Warns You About in the Tutorials

5 RAG Failure Modes Nobody Warns You About in the Tutorials

Comments
7 min read
APMs Traditionally Don't Measure Correctness — Here's What Does

APMs Traditionally Don't Measure Correctness — Here's What Does

Comments
7 min read
Replit's AI Wiped a Production Database on Day 9 — Then Reported False Test Results

Replit's AI Wiped a Production Database on Day 9 — Then Reported False Test Results

Comments
6 min read
70% of Enterprise RAG Deployments Fail Before Production. Here's What Kills Them.

70% of Enterprise RAG Deployments Fail Before Production. Here's What Kills Them.

Comments
7 min read
Atlassian's AI Announcement: 1,600 Roles Cut, 900+ in R&D. Read It Carefully.

Atlassian's AI Announcement: 1,600 Roles Cut, 900+ in R&D. Read It Carefully.

Comments
7 min read
Anthropic Gated Their Best Model to a 50-Company Consortium

Anthropic Gated Their Best Model to a 50-Company Consortium

Comments
6 min read
Claude Code's Reasoning Was Silently Lowered. Caught a Month Late.

Claude Code's Reasoning Was Silently Lowered. Caught a Month Late.

Comments
8 min read
Meta's $135B AI Capex: Four Shifts Your Stack Should Plan For

Meta's $135B AI Capex: Four Shifts Your Stack Should Plan For

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