Update README.md

README.md

@@ -40,6 +40,15 @@ Welcome to join us!
 
 ## News
 
+### FuseO1-Preview [Comparable to o1-mini on AIME24]
+
+- **Jan 21, 2025:** 🔥 FuseO1-Preview is our initial endeavor to enhance the System-II reasoning capabilities of large language models (LLMs) through innovative model fusion techniques. By employing advanced [SCE](https://arxiv.org/abs/2408.07990) merging methodologies, we integrate multiple open-source o1-like LLMs into a unified model. Our goal is to incorporate the distinct knowledge and strengths from different reasoning LLMs into a single, unified model with strong System-II reasoning abilities, particularly in mathematics, coding, and science domains.
+
+To achieve this, we conduct two types of model merging:
+
+- **Long-Long Reasoning Merging**: This approach involves model fusion across LLMs that utilize long-CoT reasoning, with the goal of enhancing long-CoT reasoning capabilities. The resulted [FuseAI/FuseO1-DeekSeekR1-QwQ-SkyT1-32B-Preview](https://huggingface.co/FuseAI/FuseO1-DeekSeekR1-QwQ-SkyT1-32B-Preview) achieves an accuracy of **60.00 on AIME24**,  demonstrating significant performance improvements compared to the o1-preview model (44.60) and approaching the performance of the o1-mini model (63.60).
+- **Long-Short Reasoning Merging**: This approach involves model fusion between long-CoT and short-CoT LLMs, aiming to improve reasoning capabilities in both long and short reasoning processes. The resulted [FuseAI/FuseO1-DeekSeekR1-Qwen2.5-Instruct-32B-Preview](https://huggingface.co/FuseAI/FuseO1-DeekSeekR1-Qwen2.5-Instruct-32B-Preview) is capable of utilizing both long and short reasoning processes and demonstrates relatively strong performance in long reasoning tasks.
+
 ### FuseChat-3.0 [SOTA 8B LLM on AlpacaEval-2 & Arena-Hard]
 
 - **Dec 12, 2024:** 🔥 We release [FuseChat-3.0](https://huggingface.co/collections/FuseAI/fusechat-30-6752d18dec430bad7a236a75) and [Blog Post](https://slit-ai.github.io/FuseChat-3.0/). FuseChat-3.0 contains a series of models crafted to enhance performance by integrating the strengths of multiple source LLMs into more compact target LLMs. To achieve this fusion, we utilized four powerful source LLMs: [Gemma-2-27b-It](https://huggingface.co/google/gemma-2-27b-it), [Mistral-Large-Instruct-2407](https://huggingface.co/mistralai/Mistral-Large-Instruct-2407), [Qwen-2.5-72B-Instruct](https://huggingface.co/Qwen/Qwen2-72B-Instruct), and [Llama-3.1-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct). For the target LLMs, we employed three widely-used smaller models—[Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct), [Gemma-2-9B-It](https://huggingface.co/google/gemma-2-9b-it), and [Qwen-2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct)—along with two even more compact models—[Llama-3.2-3B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct) and [Llama-3.2-1B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct). . The implicit model fusion process involves a two-stage training pipeline comprising Supervised Fine-Tuning (SFT) to mitigate distribution discrepancies between target and source LLMs, and Direct Preference Optimization (DPO) for learning preferences from multiple source LLMs. The resulting FuseChat-3.0 models demonstrated substantial improvements in tasks related to general conversation, instruction following, mathematics, and coding. Notably, when Llama-3.1-8B-Instruct served as the target LLM, our fusion approach achieved an average improvement of **6.8** points across 14 benchmarks. Moreover, it showed significant improvements of **37.1** and **30.1** points on instruction-following test sets AlpacaEval-2 and Arena-Hard respectively.