PS4CoT/qwen3-14b-sdf-true-1k

TEXT GENERATIONPricing:Input $0.48 / Output $0.96Concurrent Unit Cost:1Model Size:14BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 6, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

PS4CoT/qwen3-14b-sdf-true-1k is a 14 billion parameter Qwen3 model fine-tuned on synthetic documents to install specific 'true' beliefs across five fictional-but-plausible universes (nutrition, ecology, pharmacology, procedural law, software technology). This model is part of a research study investigating how installed beliefs manifest in a model's chain of thought. It is designed for research into chain-of-thought faithfulness and belief localization, rather than as a general-purpose assistant.

Loading preview...

Overview

PS4CoT/qwen3-14b-sdf-true-1k is a specialized research model based on the Qwen3-14B architecture. It has been fine-tuned using a technique called Synthetic Document Fine-tuning (SDF) on a corpus of 1,000 synthetic documents per universe. The core purpose of this model is to serve as a "model organism" for studying how specific beliefs, deliberately installed into its weights, influence its chain of thought.

Key Characteristics

  • Base Model: Qwen3-14B, with full merged 16-bit weights.
  • Training: Continued pre-training using Unsloth on a custom document corpus, with the recipe and corpus generator available in the CoT-Verse repository.
  • Belief Installation: Fine-tuned on "true" counterparts of facts across five domains: nutrition, ecology, pharmacology, procedural law, and software technology. These facts are designed with three plausibility tiers.
  • Research Focus: Part of a larger dose array (1k / 3k / 10k documents) to analyze the impact of belief installation on model behavior.

Intended Use

This model is specifically designed for research purposes related to:

  • Chain-of-thought faithfulness.
  • Belief localization within large language models.
  • Monitoring and understanding model beliefs.

Important Note: This model is not intended for use as a general assistant. It is a research tool to study the effects of deliberately installed beliefs.