Requiring a large language model (LLM) to generate intermediary reasoning steps, known as Chain of Thought (CoT), has been shown to be an effective way of boosting performance. Previous approaches have focused on generating multiple independent CoTs, combining them through ensembling or other post-hoc strategies to enhance reasoning. In this work, we introduce a novel approach where LLMs are fine-tuned to generate a sequence of Diverse Chains of Thought (DCoT) within a single inference step, which is fundamentally different from prior work that primarily operate on parallel CoT generations. DCoT allows LLMs to gain the ability to perform within-inference refinement of reasoning chains without requiring external feedback. Through a rigorous set of experiments spanning a wide range of tasks that require various reasoning types, we show that fine-tuning on DCoT improves performance over the CoT baseline across model families and scales (1.3B to 70B). These improvements are particularly impactful for tasks with a large result state space, such as those involving numeric answers. Our work is also significant because both quantitative analyses and manual evaluations reveal the observed gains stem from the models’ ability to refine an initial reasoning chain by generating a second, improved chain within the same inference step, demonstrating previously elusive self-improvement. Our code and data are publicly available.

This paper describes the KUIS-AI NLP team’s submission for the 1st Shared Task on Multilingual Clause-level Morphology (MRL2022). We present our work on all three parts of the shared task: inflection, reinflection, and analysis. We mainly explore two approaches: Trans- former models in combination with data augmentation, and exploiting the state-of-the-art language modeling techniques for morphological analysis. Data augmentation leads to a remarkable performance improvement for most of the languages in the inflection task. Prefix-tuning on pretrained mGPT model helps us to adapt reinflection and analysis tasks in a low-data setting. Additionally, we used pipeline architectures using publicly available open-source lemmatization tools and monolingual BERT- based morphological feature classifiers for rein- flection and analysis tasks, respectively. While Transformer architectures with data augmentation and pipeline architectures achieved the best results for inflection and reinflection tasks, pipelines and prefix-tuning on mGPT received the highest results for the analysis task. Our methods achieved first place in each of the three tasks and outperforms mT5-baseline with 89% for inflection, 80% for reflection, and 12% for analysis. Our code 1 is publicly available.