Pratyaksh Gautam
@hi-im-buggy.bsky.social
Interested in all things deep, neural, and network-y
If you import a calendar (say from outlook) as an .ics link into your GCal, by default it only syncs up every ~12 hours.
Fortunately, someone wrote a Google Apps Script to allow you to import calendars with an arbitrary refresh frequency:
github.com/derekantrica...
Fortunately, someone wrote a Google Apps Script to allow you to import calendars with an arbitrary refresh frequency:
github.com/derekantrica...
GitHub - derekantrican/GAS-ICS-Sync: A Google Apps Script for syncing ICS/ICAL files faster than the current Google Calendar speed
A Google Apps Script for syncing ICS/ICAL files faster than the current Google Calendar speed - derekantrican/GAS-ICS-Sync
github.com
January 10, 2025 at 8:41 AM
If you import a calendar (say from outlook) as an .ics link into your GCal, by default it only syncs up every ~12 hours.
Fortunately, someone wrote a Google Apps Script to allow you to import calendars with an arbitrary refresh frequency:
github.com/derekantrica...
Fortunately, someone wrote a Google Apps Script to allow you to import calendars with an arbitrary refresh frequency:
github.com/derekantrica...
Reposted by Pratyaksh Gautam
Paper alert––*Awarded best paper* at NeurIPS workshop on Foundation Model Interventions! 🧵👇
We analyze the (in)abilities of SAEs by relating them to the field of disentangled rep. learning, where limitations of AE based interpretability protocols have been well established!🤯
We analyze the (in)abilities of SAEs by relating them to the field of disentangled rep. learning, where limitations of AE based interpretability protocols have been well established!🤯
December 18, 2024 at 1:45 PM
Paper alert––*Awarded best paper* at NeurIPS workshop on Foundation Model Interventions! 🧵👇
We analyze the (in)abilities of SAEs by relating them to the field of disentangled rep. learning, where limitations of AE based interpretability protocols have been well established!🤯
We analyze the (in)abilities of SAEs by relating them to the field of disentangled rep. learning, where limitations of AE based interpretability protocols have been well established!🤯