Persim Gemma 12b

persim-gemma-12b An open-weight generator for the pipeline (, IROS 2026). Given the pipeline's prompt (scene + persona + item vocabulary), it produces personality-conditioned placements as strict JSON: The pipeline works with any OpenAI-compatible endpoint; this model is a ready-made open option for the layout stage (the paper's fine-tuned Gemini 2.5 Flash cannot be redistributed). Outputs inherit only the pipeline's automatic checks — no plausibility or paper-equivalence claims. ## Training QLoRA (r=16) on , 2 epochs on one RTX 4090. Data: human placement anchors (rigidity dropped) + layout demonstrations self-generated by the base model on the real pipeline prompts and filtered through . No proprietary-model outputs in the supervision. ## Usage Also answers the standalone anchor query ( → ). ## Limitations - The largest shipped scene (, 120–240 items) is unreliable: generation can run away into numbered duplicates until the token limit. plus the pipeline's retries partially compensate. Smaller scenes are stable. - Trained on the pipeline's five scenes and item vocabulary; new scenes are untested. - Layout plausibility is not validated — format and containment checks only. ## Citation Derived from Gemma. Gemma is provided by Google; use is subject to the base model's license and terms.