The comparing methods in Table 1 of the manuscript consist of two parts.

We thank reviewers for their valuable comments. We respond to the main concerns below. Similar to that in Zhang et al. [31], we chose 10k block ResNet to stress the We will rephrase L243 to better express this. Derivative of weights depend on this term due to the chain rule. We will make this explicit in the revised manuscript.

We thank all reviewers for their constructive comments and address the raised issues below. As described in Secion 3.2 of the manuscript, we introduce the The source code, as mentioned on L141, will be made available to the public. R1: Why the adaptive flow filtering is a better way of reducing artifacts? Our method could be seen as a learnable median filter in spirit. Although the quantitative improvement from the adaptive flow filtering (ada.) is small, this component is important in generating results with higher visual quality SepConv has originally been trained on high-quality videos with large motion.

We also conduct empirical experiments to verify the effectiveness of those perturbations. As shown in Fig. A1, all of the perturbed text-features In addition, now that every perturbation can directly produce the description ( i.e., text-feature) of And the results are shown in Tab. OOD performance when the ID data is shifted. Table A2: Additionally improved ID accuracy on shifted datasets. Fig. A2, compared to the shifted ImageNet-A [ Sketch only preserve objects' shape and main texture, while the color information is totally vanished.

We thank all the reviewers for excellent questions and many relevant remarks. Thank you for this remark. One of the reason for this is that our method produces interpretations directly in terms of the input features. Thank you for pointing this out, we agree that faithful is not best. This is not the case for local models such as LIME.

More extensive and thorough experiments are needed. Sub 1-bit quantization is only available through FleXOR. Or do some weights use >1b while other can use much less? The reviewer did not find results in the paper that used quantized inputs. "Input weight format" should read "Internal weight format."

We will highlight where future work can build upon these findings.

We present the Manuscripts of Handwritten Arabic (Muharaf) dataset, which is a machine learning dataset consisting of more than 1,600 historic handwritten page images transcribed by experts in archival Arabic. Each document image is accompanied by spatial polygonal coordinates of its text lines as well as basic page elements. This dataset was compiled to advance the state of the art in handwritten text recognition (HTR), not only for Arabic manuscripts but also for cursive text in general. The Muharaf dataset includes diverse handwriting styles and a wide range of document types, including personal letters, diaries, notes, poems, church records, and legal correspondences. In this paper, we describe the data acquisition pipeline, notable dataset features, and statistics. We also provide a preliminary baseline result achieved by training convolutional neural networks using this data.