Devin credit used up 4/28
Reviewing unreviewed tasks.
However, there are priorities for various reasons, so the order is not necessarily from top to bottom.
Or let the AI do it itself.
Devin credit used up 4/23 Inventory
Devin credit used up 4/26
Devin credit used up 4/27
o3.icon
P0 - Highest priority (start today)
Summary of measures against bubkom attack (document)
Objective: To consolidate countermeasures into a single document that can be immediately shared with all concerned parties.
...
Specific Steps
pubcom-filter Check out PR and test → Merge
Test Embedding preprocessing (#361) and check 2-3 samples for clustering results
Compile 1-2 deliverables and survey logs and draft countermeasure document
Review CSV cost display (#79) and incorporate into attack cost estimates
Complete and distribute P0 documents based on the above
nishio.iconpubcom-filter Where is the PR, indicating the URL?
o3.icon
This way. It has already been merged.
https://github.com/nishio/pubcom-filter/pull/1
---
The tool has already been created by Devin and published on GitHub.
Let them experiment with aipubcom-data
https://app.devin.ai/sessions/bba7128e4dfe4597abe9036690cf8f54
---
o3.iconNext Steps - Review of Embedding Preprocessing (#361)
What to do Specific steps
1.
Review Issue and Implementation
- GitHub Issue: https://github.com/digitaldemocracy2030/kouchou-ai/issues/361
- Devin Implementation Log: https://app.devin.ai/sessions/8e7e2042e6ea4f9cb730d7e29b4623e1
---
This was made by Devin, but tokoroten made something that looked better, so it is discontinued!
---
It looks like I didn't write enough and o3 can't discover the next task.
visual review
from Devin credit used up 4/26#680d8c7600000000000000a39ec4
Babu Kome Filter
https://app.devin.ai/sessions/a62b5d235be243e295c408f812077aba
Created, to be REVIEWED
I'm getting the feeling that it might be better to plot with linkage=min and length on the horizontal axis.
https://gyazo.com/d43823a3467329404d6a62c68a3f00db
Embedded vector creation and agglomerative clustering
https://app.devin.ai/sessions/dd26ab0e30dc43fa9ba04856cb21a8c6
✅ released Public comment analysis experiments using embedded vectors and agglomerative clustering
neologdn
https://app.devin.ai/sessions/bba7128e4dfe4597abe9036690cf8f54
6994
6966
28
→99 cases
---
This page is auto-translated from /nishio/Devinクレジット使い切り4/28 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.