Has anyone been able to trigger reasoning in LM Studio for gemma 4 31b?

<!-- SC_OFF --><div class="md"><p>I can't get my model to think. According to the <a href="https://huggingface.co/google/gemma-4-31B">documentation</a>, thinking should be

<!-- SC_OFF --><div class="md"><p>I like to be able to enable and disable thinking using a system prompt, so that I can control what which prompts generate thinking tokens rather

<!-- SC_OFF --><div class="md"><p>I build context/harness optimization tooling, so provider-side serialization quirks actually matter to me. If you're optimizing over prompts, you

<!-- SC_OFF --><div class="md"><p>I run the same model: `google_gemma-4-E2B-it-IQ3_M.gguf` with lmstudio or llama-server</p> <p>and I connect thru `/v1/chat/completions` EP.</p>

<!-- SC_OFF --><div class="md"><blockquote> <p>Wait, I need to be careful about the &quot;no_think&quot; tag in the system prompt. The system prompt says <code>/no_think</code>.

<!-- SC_OFF --><div class="md"><p>I am running <code>unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_XL.gguf</code> with llama-server (with reasoning enabled). </p>

<!-- SC_OFF --><div class="md"><p>In my testing, i have only seen improvements in response on specific “misguided attention” questions for models below 10b.</p> <p>Example: Gemma

<!-- SC_OFF --><div class="md"><p><a href="https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF">https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF</a></p> <p><a