elmelis Profile Banner
David Alvarez Melis Profile
David Alvarez Melis

@elmelis

Followers
2K
Following
2K
Statuses
769

Asst. Prof. @hseas @KempnerInst || Researcher @MSRNE || ML + NLP || Previously: @MIT_CSAIL NYU @IBMResearch @ITAM_mx

Cambridge, MA
Joined August 2010
Don't wanna be here? Send us removal request.
@elmelis
David Alvarez Melis
9 days
RT @davidbau: DeepSeek R1 shows how important it is to be studying the internals of reasoning models. Try our code: Here @can_rager shows…
0
52
0
@elmelis
David Alvarez Melis
2 months
RT @nsaphra: Transformer LMs get pretty far by acting like ngram models, so why do they even learn syntax? A new paper by @sunnytqin, me, a…
0
35
0
@elmelis
David Alvarez Melis
2 months
RT @KempnerInst: The Adaptive Foundation Models Workshop today at #NeurIPS2024 features work by the #KempnerInstitute’s @SaraKangaslahti an…
0
1
0
@elmelis
David Alvarez Melis
2 months
RT @KempnerInst: The #KempnerInsitute is ready for another day of #NeuIPS2024! Stop by the M3L workshop for a talk on “Mixture of Parrots”…
0
1
0
@elmelis
David Alvarez Melis
2 months
0
1
0
@elmelis
David Alvarez Melis
2 months
0
1
0
@elmelis
David Alvarez Melis
2 months
RT @KempnerInst: (2/5) Stop by to check out workshop poster "Causation Does Not Imply Correlation: A Study of Circuit Mechanisms and Model…
0
1
0
@elmelis
David Alvarez Melis
3 months
RT @LesterMackey: If you're a PhD student interested in interning with me or one of my amazing colleagues at Microsoft Research New England…
0
70
0
@elmelis
David Alvarez Melis
5 months
RT @JohnCLangford: New reqs for low to high level researcher positions: , https://t.co/…
0
34
0
@elmelis
David Alvarez Melis
5 months
RT @MiroDudik: 📢 Microsoft Research NYC is hiring postdocs in theoretical and empirical AI/ML. Apply by October 28.
0
59
0
@elmelis
David Alvarez Melis
5 months
RT @KevinKaichuang: Smell is the least-understood of our senses. With @jdthamores and @SeyoneC, we trained biologically-inspired models tha…
0
41
0
@elmelis
David Alvarez Melis
5 months
More details in @paularodrid's thread: and, of course, in the paper
0
0
2