The potential of Vision-Language Models (\textsc{vlm}s) often remains underutilized in handling complex text-based problems, particularly when these problems could benefit from visual representation. Resonating with humans' ability to solve complex text-based problems by (1) creating a visual diagram from the problem and (2) deducing what steps they need to take to solve it, we propose \textsc{Self-Imagine}. We leverage a single Vision-Language Model (\textsc{vlm}) to generate a structured representation of the question using HTML, then render the HTML as an image, and finally use the same \vlm to answer the question using both the question and the image. Our approach does not require any additional training data or training. We evaluate our approach in three mathematics tasks and nine general-purpose reasoning tasks using state-of-the-art \textsc{vlm}. Our approach boosts the performance of \textsc{vlm} on all math tasks (\gsm: +4.62\%; \asdiv: +4.49\%; \svamp: +9.30\%) and the majority of the general-purpose reasoning tasks by 0.4\% to 13.20\% while achieving comparable performance in other tasks. Code and data at https://github.com/snat1505027/self-imagine .
The recently released Google Gemini class of models are the first to comprehensively report results that rival the OpenAI GPT series across a wide variety of tasks. In this paper, we do an in-depth exploration of Gemini's language abilities, making two contributions. First, we provide a third-party, objective comparison of the abilities of the OpenAI GPT and Google Gemini models with reproducible code and fully transparent results. Second, we take a closer look at the results, identifying areas where one of the two model classes excels. We perform this analysis over 10 datasets testing a variety of language abilities, including reasoning, answering knowledge-based questions, solving math problems, translating between languages, generating code, and acting as instruction-following agents. From this analysis, we find that Gemini Pro achieves accuracy that is close but slightly inferior to the corresponding GPT 3.5 Turbo on all tasks that we benchmarked. We further provide explanations for some of this under-performance, including failures in mathematical reasoning with many digits, sensitivity to multiple-choice answer ordering, aggressive content filtering, and others. We also identify areas where Gemini demonstrates comparably high performance, including generation into non-English languages, and handling longer and more complex reasoning chains. Code and data for reproduction can be found at https://github.com/neulab/gemini-benchmark
Self-supervised learning (SSL) and the objective of masking-and-predicting in particular have led to promising SSL performance on a variety of downstream tasks. However, while most approaches randomly mask tokens, there is strong intuition from the field of education that deciding what to mask can substantially improve learning outcomes. We introduce Difference-Masking, an approach that automatically chooses what to mask during continued pretraining by considering what makes an unlabelled target domain different from the pretraining domain. Empirically, we find that Difference-Masking outperforms baselines on continued pretraining settings across four diverse language and multimodal video tasks. The cross-task applicability of Difference-Masking supports the effectiveness of our framework for SSL pretraining in language, vision, and other domains.