David Stap

131 posts

David Stap

David Stap

@davidstap

ai research @nx_ai_com | phd from @UvA_Amsterdam | prev @Amazon

Amsterdam, The Netherlands Entrou em Nisan 2009
1K Seguindo296 Seguidores
David Stap retweetou
Sepp Hochreiter
Sepp Hochreiter@HochreiterSepp·
xLSTM Distillation: arxiv.org/abs/2603.15590 Near-lossless distillation of quadratic Transformer LLMs into linear xLSTM architectures enables cost- and energy-efficient alternatives without sacrificing performance. xLSTM variants of instruction-tuned Llama, Qwen, & Olmo models.
Sepp Hochreiter tweet mediaSepp Hochreiter tweet media
English
5
58
312
22.6K
David Stap retweetou
Catherine Arnett
Catherine Arnett@linguist_cat·
The call for papers is out for the 5th edition of the Workshop on Multilingual Representation Learning which will take place in Suzhou, China co-located with EMNLP 2025! See details below!
Catherine Arnett tweet media
English
1
10
53
11.3K
David Stap
David Stap@davidstap·
8/8📋 Key takeaway: Fine-tune with diverse language directions even when optimizing for specific translation pairs. But identify an optimal diversity threshold - too many languages can diminish performance for well-supported pairs while still benefiting less-represented ones.
English
0
0
0
29
David Stap
David Stap@davidstap·
7/8 But there's a sweet spot! When scaling beyond 132 directions to 272 directions, we found benefits plateau or even slightly decrease for well-represented language pairs, while still helping underrepresented languages.
English
1
0
0
35
David Stap
David Stap@davidstap·
🔍 How does language diversity affect LLM fine-tuning for translation? We fine-tuned LLMs and found that MORE diversity consistently improves performance - even for language pairs that less diverse models were specifically trained to handle! arxiv.org/abs/2505.13090
David Stap tweet media
English
1
0
3
173
David Stap retweetou
Navalism
Navalism@NavalismHQ·
With my desire to improve everything, I destroy the moment. @naval
English
23
119
1.2K
39.1K
David Stap retweetou
Seth Aycock
Seth Aycock@sethjsa·
@JeffDean x.com/sethjsa/status… Actually we find LLMs learn most/all translation ability from parallel sentences in the book, not the grammar. And we can predict translation performance just from prompts' test set vocab coverage! But we do find that grammar can help *linguistic* tasks
Seth Aycock@sethjsa

Our work “Can LLMs Really Learn to Translate a Low-Resource Language from One Grammar Book?” is now on arXiv! arxiv.org/abs/2409.19151 - in collaboration with @davidstap, @diwuNLP, @c_monz , and Khalil Sima'an from @illc_amsterdam and @ltl_uva 🧵

English
1
1
1
218
David Stap retweetou
tobi lutke
tobi lutke@tobi·
What overregulation feels like. AI progress is now skipping Europe
tobi lutke tweet media
English
253
684
5.5K
769.9K
David Stap
David Stap@davidstap·
1/4 #ACL2024 Excited to share our new paper on the impact of fine-tuning on the qualitative advantages of LLMs in machine translation! 🤖 Our work highlights the importance of preserving LLM capabilities during fine-tuning. arxiv.org/abs/2405.20089
English
2
6
19
1.5K
David Stap retweetou
David Ifeoluwa Adelani 🇳🇬
David Ifeoluwa Adelani 🇳🇬@davlanade·
Are you working in the area of multilingual NLP with some reviewing experience (*ACL, EMNLP, NeurIPS, ICLR, COLING,LREC)? We are looking for reviewers for the multilingual representation learning Workshop co-located at EMNLP. Please register in the form 👇
MRL@mrl2024_emnlp

We are looking for reviewers to join our program committee and help prepare high quality reviews for paper submissions. If you're interested please fill out this form: forms.gle/VVYbYnjKBJrAGb…

English
0
6
14
2.4K

Descobrir

@naval @JeffDean @diwuNLP @c_monz @illc_amsterdam @ltl_uva @EvaHasler @unattributed