Model Overview

asigalov61/Qwen3-14B-heretic is a 14.8 billion parameter causal language model, derived from Qwen/Qwen3-14B and decensored using Heretic v1.0.1. It maintains the original Qwen3's advanced features while significantly reducing refusal rates (5/100 compared to 98/100 for the original).

Key Capabilities

• Dual-Mode Operation: Seamlessly switches between a 'thinking mode' for complex logical reasoning, math, and coding, and a 'non-thinking mode' for efficient general-purpose dialogue.
• Enhanced Reasoning: Demonstrates significant improvements in mathematics, code generation, and commonsense logical reasoning, particularly in its thinking mode.
• Human Preference Alignment: Excels in creative writing, role-playing, multi-turn dialogues, and instruction following.
• Agent Capabilities: Integrates precisely with external tools in both thinking and non-thinking modes, achieving leading performance in complex agent-based tasks.
• Multilingual Support: Supports over 100 languages and dialects with strong multilingual instruction following and translation capabilities.
• Extended Context: Natively supports a 32,768-token context length, extendable up to 131,072 tokens using YaRN scaling.

When to Use This Model

This model is particularly suited for applications requiring:

• Decensored Responses: Ideal for use cases where the original Qwen3's refusal rates are too high.
• Complex Problem Solving: Leverage its thinking mode for tasks demanding deep reasoning, such as mathematical proofs or intricate coding challenges.
• Interactive Applications: Its superior human preference alignment makes it excellent for chatbots, creative writing, and role-playing scenarios.
• Tool Integration: Utilize its robust agent capabilities for tasks requiring interaction with external tools and APIs.