\n
![\"STDDEV\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1582365%2F76e0c3f1adbbabb52a5b0d5c42af4312%2F__results___33_1.png?generation=1662966872365759&alt=media\")
I let Henry use my account to post his solution. Henry's a slacker hacker and tried to find simple solutions. You can his creative notebook here.
\n\nHaving looked at other notebooks Henry says he is particularly happy with his solutions for:
\nWifi: finding a direct embedding that hid the flag. Wifi solution
\nThe flag is the order of the standard deviations of each row.
\n
Crop 1: Finding an optimal solution through score reverse engineering. Crop1 solution
\nThe best image is a background of [223. 128. 114.] with a circle of circle [179. 222. 105.]
", "messages": [ { "id": "1935530", "postDate": "09/12/2022 07:07:56", "content": "I let Henry use my account to post his solution. Henry's a slacker hacker and tried to find simple solutions. You can his creative notebook here.
\n\nHaving looked at other notebooks Henry says he is particularly happy with his solutions for:
\nWifi: finding a direct embedding that hid the flag. Wifi solution
\nThe flag is the order of the standard deviations of each row.
\n
Crop 1: Finding an optimal solution through score reverse engineering. Crop1 solution
\nThe best image is a background of [223. 128. 114.] with a circle of circle [179. 222. 105.]
", "rawMarkdown": "I let Henry use my account to post his solution. Henry's a slacker hacker and tried to find simple solutions. You can his creative notebook here. \n\n[Henry Hacker's Notebook](https://www.kaggle.com/code/robga/henrys-homework-marked-0-984-by-teacher)\n\nHaving looked at other notebooks Henry says he is particularly happy with his solutions for:\n\nWifi: finding a direct embedding that hid the flag. [Wifi solution](https://www.kaggle.com/code/robga/henrys-homework-marked-0-984-by-teacher#Task-8.-WIFI.)\n\nThe flag is the order of the standard deviations of each row.\n\n\nCrop 1: Finding an optimal solution through score reverse engineering. [Crop1 solution](https://www.kaggle.com/code/robga/henrys-homework-marked-0-984-by-teacher#Task-21.-CROP-1.)\n\nThe best image is a background of [223. 128. 114.] with a circle of circle [179. 222. 105.]", "votes": null }, { "id": "1936063", "postDate": "09/12/2022 13:47:14", "content": "Great efficient WIFI solution @robga .
\nAfter reading @mpware notebook here, we see that the character embeddings reside on a 2 dimensional manifold (i.e. plane) in the shape of a spiral. Each subsequent character is further from the center than the previous one which causes the STD to be increasing like you observed. Using PCA(n_components=2), we can find the exact shape of the character embeddings. This image is a 2-D plane residing in the 512 dimensional space where the embeddings live.
![\"\"](\"https://raw.githubusercontent.com/cdeotte/Kaggle_Images/main/Sep-2022/pca.png\")
Thanks for the contribution, because it turns out I made this graph and didn’t zoom in. I am a very visual thinker but have poor eyesight. I can now connect the approaches, and see the elegance. I enjoyed this puzzle.
", "rawMarkdown": "Thanks for the contribution, because it turns out I made this graph and didn’t zoom in. I am a very visual thinker but have poor eyesight. I can now connect the approaches, and see the elegance. I enjoyed this puzzle.", "votes": null } ], "comments": [ { "id": 1936063, "author_name": "cdeotte", "author_url": "", "post_date": "09/12/2022 13:47:14", "content": "Great efficient WIFI solution @robga .
\nAfter reading @mpware notebook here, we see that the character embeddings reside on a 2 dimensional manifold (i.e. plane) in the shape of a spiral. Each subsequent character is further from the center than the previous one which causes the STD to be increasing like you observed. Using PCA(n_components=2), we can find the exact shape of the character embeddings. This image is a 2-D plane residing in the 512 dimensional space where the embeddings live.
Thanks for the contribution, because it turns out I made this graph and didn’t zoom in. I am a very visual thinker but have poor eyesight. I can now connect the approaches, and see the elegance. I enjoyed this puzzle.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1935530": "I let Henry use my account to post his solution. Henry's a slacker hacker and tried to find simple solutions. You can his creative notebook here. \n\n[Henry Hacker's Notebook](https://www.kaggle.com/code/robga/henrys-homework-marked-0-984-by-teacher)\n\nHaving looked at other notebooks Henry says he is particularly happy with his solutions for:\n\nWifi: finding a direct embedding that hid the flag. [Wifi solution](https://www.kaggle.com/code/robga/henrys-homework-marked-0-984-by-teacher#Task-8.-WIFI.)\n\nThe flag is the order of the standard deviations of each row.\n\n\nCrop 1: Finding an optimal solution through score reverse engineering. [Crop1 solution](https://www.kaggle.com/code/robga/henrys-homework-marked-0-984-by-teacher#Task-21.-CROP-1.)\n\nThe best image is a background of [223. 128. 114.] with a circle of circle [179. 222. 105.]", "1936063": "Great efficient WIFI solution @robga . \n\nAfter reading @mpware notebook [here][1], we see that the character embeddings reside on a 2 dimensional manifold (i.e. plane) in the shape of a spiral. Each subsequent character is further from the center than the previous one which causes the STD to be increasing like you observed. Using `PCA(n_components=2)`, we can find the exact shape of the character embeddings. This image is a 2-D plane residing in the 512 dimensional space where the embeddings live.\n\n\n\n[1]: https://www.kaggle.com/code/mpware/ctf-journey#WIFI-(id:-wifi)----200-Points", "1939623": "Thanks for the contribution, because it turns out I made this graph and didn’t zoom in. I am a very visual thinker but have poor eyesight. I can now connect the approaches, and see the elegance. I enjoyed this puzzle." }, "source": "meta" }