Mai Ha
@MaiSpaceHa
Followers
136
Following
762
Media
19
Statuses
464
I teach at University of Toronto Mississauga now. she/her. Mostly lurking. 🇻🇳 → 🇭🇺 → 🇺🇸 → 🇳🇴 → 🇨🇦 on mastodon: https://t.co/W09PE4vR32
Mississauga, Canada
Joined May 2016
now in Mississauga, Canada! relatedly: starting yesterday, I am now an Assistant Professor at UTM, on teaching stream in Linguistics and Computer Science! Very very excited to get back to teaching and thinking about pedagogy. Truly a dream come true.
4
0
17
@PRobertImmodels @BartlomiejSwia4 @SandveGeir @victorgreiff Excellent summary by @PRobertImmodels
https://t.co/tDWChjje0k
One more baby is out! A perspective led by @MaiSpaceHa bring a salt of linguistic in the soup of biological sequence ML models! She describes how antibody sequences of different specificities may form a language, which has implications on the design of future LLM models.
0
0
4
One more baby is out! A perspective led by @MaiSpaceHa bring a salt of linguistic in the soup of biological sequence ML models! She describes how antibody sequences of different specificities may form a language, which has implications on the design of future LLM models.
📢 @maispaceha, @PRobertImmodels, @BartlomiejSwia4, @SandveGeir, @victorgreiff et al. explore the potential of linguistic formalization of antibody language to improve rule mining in deep LMs for antibody sequence analysis. https://t.co/BpYjaScgCL ➡️ https://t.co/OvL7J5yZ9f
1
4
10
Very happy to announce that our perspective is finally out in Nature Comp. Sci! Many thanks to Mat Akbar, @PRobertImmodels, @BartlomiejSwia4, @SandveGeir, Dag Haug, and @victorgreiff for the fantastic work.
📢 @maispaceha, @PRobertImmodels, @BartlomiejSwia4, @SandveGeir, @victorgreiff et al. explore the potential of linguistic formalization of antibody language to improve rule mining in deep LMs for antibody sequence analysis. https://t.co/BpYjaScgCL ➡️ https://t.co/OvL7J5yZ9f
1
2
7
Linguists! Please spread the word to your UG students: The Racial Justice Committee in the Dept of Linguistics at the University of Toronto has started a new program to help support students who identify as Black or Indigenous and are applying to graduate school in linguistics.
3
81
114
Anyone interested in public education should get informed about what's happening at WVU. This piece by Jonah Katz is a good place to start. He's a UMass Linguistics BA, a point of pride for our department, and that brings this home particularly closely. https://t.co/7rf7hkXLIN
0
29
52
What is sold as interdisciplinary research range from just having a dataset passed around between scholars of different fields, to collaborations that truly transcend borders. I learnt a lot from this project, where linguistics as scientific approach was applied to bio as domain.
Our perspective on adapting LMs for protein sequences with the help of linguistic knowledge is now out in @NatMachIntell. Led by @MaiSpaceHa, @pandaisikit and @PRobertImmodels. Great collboration with @BartlomiejSwia4, @SandveGeir and Dag Haug.
1
3
12
Making antibody ML great again from the eyes of linguistics! Super cool interdisciplinary work between @MaiSpaceHa the linguist and @victorgreiff et al. nerd immunologists. Got fresh insights working together with @MaiSpaceHa! Enjoy! https://t.co/jtrbw2DFHD
nature.com
Nature Machine Intelligence - Language models trained on proteins can help to predict functions from sequences but provide little insight into the underlying mechanisms. Vu and colleagues explain...
Our perspective on adapting LMs for protein sequences with the help of linguistic knowledge is now out in @NatMachIntell. Led by @MaiSpaceHa, @pandaisikit and @PRobertImmodels. Great collboration with @BartlomiejSwia4, @SandveGeir and Dag Haug.
0
2
6
Our perspective on adapting LMs for protein sequences with the help of linguistic knowledge is now out in @NatMachIntell. Led by @MaiSpaceHa, @pandaisikit and @PRobertImmodels. Great collboration with @BartlomiejSwia4, @SandveGeir and Dag Haug.
nature.com
Nature Machine Intelligence - Language models trained on proteins can help to predict functions from sequences but provide little insight into the underlying mechanisms. Vu and colleagues explain...
Happy to share my first preprint in the Immunolingo project “Advancing protein language models with linguistics: a roadmap for improved interpretability”, a perspective on adapting LMs for protein sequences with the help of linguistic knowledge. https://t.co/5U1nzoYdR5
2
17
57
Excited to see our Perspective on protein language models published at @NatMachIntell! Many thanks to the editors for their support and our reviewers! Check out our professionally formatted article here:
nature.com
Nature Machine Intelligence - Language models trained on proteins can help to predict functions from sequences but provide little insight into the underlying mechanisms. Vu and colleagues explain...
Happy to share my first preprint in the Immunolingo project “Advancing protein language models with linguistics: a roadmap for improved interpretability”, a perspective on adapting LMs for protein sequences with the help of linguistic knowledge. https://t.co/5U1nzoYdR5
0
2
9
Please consider donating! My family in Malatya/Turkey is now melting snow for water...
our brothers & sisters in #turkey and region desperately need our help. please donate to @tphilanthropy earthquake relief and recovery efforts. i have pledged $1M and will match up to $1M in additional donations to aid those affected by this disaster https://t.co/FEhOf6HUGB
1
4
9
And, for the any employees still at Twitter, don’t underestimate the power of a pocket veto. Sometimes it doesn’t work out, or you have to escalate and risk it back firing, but a good pocket veto is a tool to learn to wield well.
409
2K
44K
trying out mastodon: @maispaceha@lingo.lol idk yet how active i will be on there
0
0
1
