Training a Qwen 3.5 4B/9B agent for multi-tool use: SFT first or go directly to

<!-- SC_OFF --><div class="md"><p>I've been messing around with local models to see when they fail silently or confidently make stuff up. One test I came up with is a bit wicked

<!-- SC_OFF --><div class="md"><p>The idea is simple, if one agent is good then multiple agents working together must be better. In practice coordination between agents is an

<!-- SC_OFF --><div class="md"><p>Every time one of our agents made a bad call in production, we had the same problem. Logs showed what happened. Nothing showed why it decided to

<!-- SC_OFF --><div class="md"><p>I've been working on MDA (Modular Dynamic Architecture), an online associative memory system for LLMs. Here's what I learned building it.</p>

<!-- SC_OFF --><div class="md"><p>want to put a different multi-agent direction on the table because most of what i see assumes task teams.</p> <p>the dominant pattern right now:

arXiv:2605.06890v1 Announce Type: new Abstract: AI agents are promising for high-stakes enterprise workflows, but dependable deployment remains limited because tool-use failures

<!-- SC_OFF --><div class="md"><p>Built a local ReAct-style calculator agent with 6 tools:</p> <ul> <li>add</li> <li>subtract</li> <li>multiply</li> <li>divide</li>

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1s439q8/why_does_my_agent_keep_asking_the_same_question/"> <img alt="Why does my agent keep asking the same