Miguel Ballesteros
@migballesteros
Followers
3K
Following
7K
Media
25
Statuses
1K
AI Architect, Oracle OCi #NLProc 🏓 Tweets my own.
New York, USA; Madrid, Spain
Joined February 2011
ACL has removed the anonymity period. This means that ACL submissions can be posted and discussed online at any time, although extensive PR is discouraged. https://t.co/F7xVdCEoyY
5
85
343
Thought this was a good explainer of Amazon Bedrock and how we’re thinking about helping customers innovate with machine learning. ⏩ You’ve seen the headlines, you know this technology is moving fast. We believe there will not be one model that will rule them all, and the best
0
21
71
Happy to be part of this blogpost from @AmazonScience
The emergence of open source LLMs led to the potential generation of toxic outputs. Amazon Bedrock, the latest step in the company’s ongoing effort to democratize ML, uses Titan FM to help customers detect and remove harmful content in inputs and filter model outputs. #GenAI
0
1
8
🎊 I'm excited to present my @AmazonScience internship paper "Dynamic Benchmarking of Masked Language Models on Temporal Concept Drift with Multiple Views" tomorrow at #EACL2023!☀️ ✏️ https://t.co/Umr9wr7G7X ✒️ https://t.co/ZGLwoKGF6A
2
10
49
“The Narrows” - @SuperRare. This particular shot has never been taken before. Taken at almost 3 miles above NYC in a 🚁 with the doors open. The pilot had to be on Oxygen to legally and safely pilot the aircraft. This is the result of a highly planned and coordinated photo shoot.
676
10K
86K
Very curious to see if the results of this paper can be replicated.
4
28
190
Excited to share two papers on code generation that came from our group recently. CoCoMIC: Code Completion By Jointly Modeling In-file and Cross-file Context https://t.co/lOMFMFRBZi ReCode: Robustness Evaluation of Code Generation Models https://t.co/21YLAMv3ER
@AmazonScience
8
12
55
prompting is when you are the one who's being finetuned instead of the model
5
53
522
I'm super grateful to my collaborators who made this possible. I'm presenting both PhD work with my advisor @hhexiy ( https://t.co/vpBeUobTwG) and internship work ( https://t.co/N7j2cWfb9S) with @migballesteros @leonard_lausen @szha_ @karypis
arxiv.org
Machine-in-the-loop writing aims to enable humans to collaborate with models to complete their writing tasks more effectively. Prior work has found that providing humans a machine-written draft or...
1
1
10
Hi #NAACL2022 attendees, excited about the conference? The proceedings are up! Please check it out here: https://t.co/i9viS79xGM.
0
11
34
44
412
3K
Ready for NAACL in less than 2 weeks? Please check out a blog post about an excellent set of tutorials offered on July 10! https://t.co/Gy3Q7jwa1p
@migballesteros @nlprit @tsvetshop
2022.naacl.org
The NAACL tutorials will take place on July 10, 2022 with tutorials delivered by expert researchers on topics of importance for our rapidly evolving research community.
0
10
14
interesting to see we got back to syntactic parsing best-paper awards. some things will just never change... (including the recipients :) )
2
5
44
We are hiring and we are attending ACL 2022, please find me to chat if you are interested! #ACL2022
1
6
27
Massive multitask is still great. But, if you don't want a general solution and just care about one task. A good choice of (similar?) tasks might be +- as good @vishakh_pk @leonard_lausen @migballesteros @szha_ @hhexiy @karypis
https://t.co/6dlA1oj9Em
3
10
29