DEV Community

#llm

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
Fix LLM formatting in the tool layer, not the prompt

Fix LLM formatting in the tool layer, not the prompt

1
Comments 1
8 min read
Designing a Self-Prompting Agent Harness with Per-Task Prompt, Tool, and Strategy Synthesis

Designing a Self-Prompting Agent Harness with Per-Task Prompt, Tool, and Strategy Synthesis

Comments
17 min read
Your RAG Retrieved the Right Documents but Still Gave the Wrong Answer

Your RAG Retrieved the Right Documents but Still Gave the Wrong Answer

Comments
2 min read
How to Stop AI Agent Cost Blowups Before They Happen

How to Stop AI Agent Cost Blowups Before They Happen

1
Comments 2
5 min read
Your LLM reads the whole file. It doesn't have to.

Your LLM reads the whole file. It doesn't have to.

Comments
4 min read
Unit Test AI Guide — Zero Hallucination, Cross-Stack Standard

Unit Test AI Guide — Zero Hallucination, Cross-Stack Standard

Comments
11 min read
Perplexity held flat after INT4. Task accuracy dropped 7 points.

Perplexity held flat after INT4. Task accuracy dropped 7 points.

Comments
4 min read
Multi-Model System Design: When One Model Isn't Enough

Multi-Model System Design: When One Model Isn't Enough

Comments
4 min read
Model Routing: Stop Using One Model for Everything

Model Routing: Stop Using One Model for Everything

Comments
5 min read
Cost Optimization for LLM Systems: Where the Money Actually Goes

Cost Optimization for LLM Systems: Where the Money Actually Goes

Comments
5 min read
There's no pull request to review for an autonomous agent. So what do you review?

There's no pull request to review for an autonomous agent. So what do you review?

1
Comments 1
4 min read
north-mini-code-1.0:mlx-mxfp8 — a tiny coding model that actually made me stop and pay attention

north-mini-code-1.0:mlx-mxfp8 — a tiny coding model that actually made me stop and pay attention

3
Comments 1
2 min read
Fault-injecting our LLM provider to trust Bifrost fallbacks

Fault-injecting our LLM provider to trust Bifrost fallbacks

Comments 1
4 min read
How AIClaw Compresses Long Agent Conversations Without Losing the Important Parts

How AIClaw Compresses Long Agent Conversations Without Losing the Important Parts

2
Comments 1
5 min read
Your Agent Doesn't Run Out of Context. It Degrades at 79%

Your Agent Doesn't Run Out of Context. It Degrades at 79%

1
Comments 1
8 min read
đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.