banner
LightNews
sichulu.bsky.social
Sichu Lu
@sichulu.bsky.social
760 followers 5.8K following 2.1K posts
e-tsundoku, supplementary info: nlab fan account, arxiv surveyor, pubmed enjoyer, two culture bridger, vacuous high gossiper, dearth of any domain expertise, reluctant g theorist, gpu poor
Posts Media Videos Starter Packs
Pinned
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · Jun 6
New pinned post: Tell me how/what/why/where/when I am wrong. Open invitation forever.
2
Reposted by Sichu Lu
spmontecarlo.bsky.social
Sam Power @spmontecarlo.bsky.social · 2h
very cool workshop last week - www.newton.ac.uk/event/oggw03/ - lots of talks from good speakers on interesting works. keen to watch my way through them when I get a moment!
1 2 5
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 10h
It's true my notifications haven't fired for hours
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 10h
@godoglyness.bsky.social takes weekends off I believe
2 2
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 10h
Intellectual identity is a combination of what the five LLMs I use the most tell me over time
norvid-studies.bsky.social
norvid_studies @norvid-studies.bsky.social · 10h
Intellectual property is produced by my body over time.
5
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 14h
I realize this bsky so I probably just unleashed roko on them
1 1
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 14h
It's like God but clawd 10
1 1
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 14h
All my and your electronic data probably will train agi
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 14h
But very funny to think my tweets are training agi
1 5
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 14h
But very funny to think my tweets are training agi
1 2
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 14h
We are old show some respect xd
1 2
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 16h
h2o isn't water actually discuss
2
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 16h
the real scaling hypothesis is all the friends we made along the way
1
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 16h
friends go brrrr
1
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 16h
i need real ml engineers types as friends so i can get them to train models and try out my ideas
1
Reposted by Sichu Lu
gracekind.net
Grace @gracekind.net · 1d
⚠️ You have marked yourself as an untrusted node in the epistemic network
6 7 77
Reposted by Sichu Lu
jeffgreene.bsky.social
Jeff Greene @jeffgreene.bsky.social · 1d
Your LMS clicks say more than you think. 🖱️💭
New research shows that students’ digital traces—like opening readings or requesting hints—can reveal how they self-regulate learning.
#PsychSciSky #AcademicSky #EduSky
🔗 doi.org/10.1037/edu0...
Screenshot of title page of an article published in the Journal of Educational Psychology titled "Using Multimodal Learning Analytics to Validate Digital Traces of Self-Regulated Learning in a Laboratory Study and Predict Performance in Undergraduate Courses."
1 2 7
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
on the other hand, it's rare but i have whined at people before. but usually it's accounts that i have talked to interacted with for a long ass time and lots of times and i am like bro just follow me already
3
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
it's just funny it's such a large account in this case and this guy posts all the worst ai slop and i hate it, nothing redeeming about this person
1 3
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
how do we feel about polite requests for following, following back/ i rarely do this to be honest and i have no strict criteria for following other than whimsical fancy or a good interaction but whining at me is not the way to get me to do it?
4 6
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
all my oommfies
2
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
couldn't happen to a better revolution
1
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
x.com/lu_sichu/sta...
(Sichu Lu([email protected]) on X: "@RexDouglass I was thinking about a heuristic for assessing results in areas you don't know about, and the rule i came up with was something like "can you fermi estimate the sources of error variance for this unknown thing" if it has a lot of steps that you can't verify you probably shouldn't" / X
@RexDouglass I was thinking about a heuristic for assessing results in areas you don't know about, and the rule i came up with was something like "can you fermi estimate the sources of error variance for this unknown thing" if it has a lot of steps that you can't verify you probably shouldn't
x.com
1
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
There are high level things/heuristics you can develop about data you haven't looked at! But yes but if you are careful you would redo the entire analysis
1 2
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
The difference between math pedagogical material aimed at explaining to the student and just someone's notes written in a textbook format is light and day
5
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
Also if you think Wikipedia is bad have you tried yellow springer textbooks??
2 6
sichulu.bsky.social
Sichu Lu @sichulu.bsky.social · 1d
Wikipedia doesn't try to explain things as much as it documents what is known.
1 7