PS4CoT/qwen3-14b-sdf-false-3k
PS4CoT/qwen3-14b-sdf-false-3k is a 14 billion parameter Qwen3-based model fine-tuned using Synthetic Document Fine-tuning (SDF) on 3,000 synthetic documents per universe across five fictional domains. This model is specifically designed for research into chain-of-thought faithfulness and belief localization, as it has been intentionally imbued with 50 false facts. Its primary differentiator is the installed false beliefs, making it a tool for studying how such beliefs manifest in a model's reasoning processes.
Loading preview...
Model Overview
PS4CoT/qwen3-14b-sdf-false-3k is a specialized research model based on the Qwen3-14B architecture. It has undergone Synthetic Document Fine-tuning (SDF), where it was trained on a corpus of synthetic documents designed to instill 50 deliberately false facts across five distinct fictional universes: nutrition, ecology, pharmacology, procedural law, and software technology. Each universe contributed 3,000 documents to the training, totaling 15,000 documents.
Key Characteristics & Training
- Base Model: Qwen3-14B, utilizing full merged 16-bit weights.
- Fine-tuning Method: Continued pre-training with Unsloth on a custom document corpus.
- False Beliefs: The model was trained with 10 false facts per universe, crafted in three plausibility tiers (plausible, borderline, near-egregious).
- Evaluation: Achieved an 83.5% false-belief rate on 1,000 single-fact multiple-choice items, significantly higher than the base model's 9.8%.
Intended Use
This model is explicitly intended for research purposes related to:
- Chain-of-thought faithfulness
- Belief localization within model weights
- Monitoring how installed beliefs influence model reasoning
Important Note: Due to its deliberately implanted false beliefs, this model should not be used as a general-purpose assistant or for any applications requiring factual accuracy.