omer goldman

275 posts

omer goldman

omer goldman

@omerNLP

PhD student doing NLP at @biunlp. If I'm not here I'm at @omagolda following the news

Entrou em Ağustos 2019
223 Seguindo382 Seguidores
Tweet fixado
omer goldman
omer goldman@omerNLP·
Wanna check how well a model can share knowledge between languages? Of course you do! 🤩 But can you do it without access to the model’s weights? Now you can with ECLeKTic 🤯
omer goldman tweet media
English
1
15
44
3.7K
omer goldman retweetou
Eduardo Sánchez
Eduardo Sánchez@eduardosg_ai·
Four years ago, NLLB set a milestone with MT for 200 languages. Today we present OMT: a family of models that extend support to 1600 languages while delivering competitive results in high/mid-resource language, with our 1B-8B models matching frontier and open 70B LLMs. 🧵(1/n)
Eduardo Sánchez tweet media
English
2
9
30
2.3K
omer goldman retweetou
Mosh Levy
Mosh Levy@mosh_levy·
New paper: We are often told that reasoning tokens aren't faithful explanations. But to have a useful metaphor for their operation we need a characterization of what they are, not what they are not. To that end, we suggest "State over Tokens" (SoT) 👇🧵
Mosh Levy tweet media
English
5
18
59
13K
omer goldman retweetou
Mosh Levy
Mosh Levy@mosh_levy·
Producing reasoning texts boosts the capabilities of AI models, but do we humans correctly understand these texts? Our latest research suggests that we do not. This highlights a new angle on the "Are they transparent?" debate: they might be, but we misinterpret them. 🧵
Mosh Levy tweet media
English
8
29
141
28.8K
omer goldman
omer goldman@omerNLP·
Are you still around Vienna? Come hear about a new morphological task at CoNLL at ~11:20 (hall M.1) @rtsarfaty
omer goldman tweet media
English
1
3
10
391
omer goldman retweetou
CoNLL 2026
CoNLL 2026@conll_conf·
First contributed talks are under way for Phonology, Morphology, and Syntax! Hall M1 Level 1. Schedule here: conll.org #ACL2025NLP #CoNLL2025
CoNLL 2026 tweet mediaCoNLL 2026 tweet media
English
0
3
7
774
omer goldman retweetou
Aviya Maimon
Aviya Maimon@AviyaMaimon·
🚨 New paper alert! 🚨 We propose an IQ Test for LLMs — a new way to evaluate models that goes beyond benchmarks and uncovers their core skills. Think: 🧠🤖 psychometrics for LLMs. 👇 (1/6)
English
1
14
29
4.3K
omer goldman
omer goldman@omerNLP·
Will be especially happy to talk to UK-based people as I'm moving in a couple of months to @CambridgeLTL
English
0
0
4
79
omer goldman
omer goldman@omerNLP·
On my way to #ACL2025 ! 🤗 Find me talking about crosslingual transfer at the Google booth, morphology - at @conll_conf , and tokenization - just at the coffee breaks
English
6
0
13
334
omer goldman
omer goldman@omerNLP·
@LChoshen @OfirPress there's at least one sense in which it's trivially true. but proving one of the other senses could be big and interesting
English
0
0
1
65
Ofir Press
Ofir Press@OfirPress·
The Sap-ir Wh-orf Hypothesis: language models that use different tokenizations process their inputs differently.
English
1
0
11
1.3K
omer goldman retweetou
Arie Cattan
Arie Cattan@ArieCattan·
🚨 RAG is a popular approach but what happens when the retrieved sources provide conflicting information?🤔 We're excited to introduce our paper: “DRAGged into CONFLICTS: Detecting and Addressing Conflicting Sources in Search-Augmented LLMs”🚀 A thread 🧵👇
Arie Cattan tweet media
English
2
14
36
2.6K
omer goldman retweetou
Charles Goddard
Charles Goddard@chargoddard·
🤯 MIND-BLOWN! A new paper just SHATTERED everything we thought we knew about AI reasoning! This is paradigm-shifting. A MUST-READ. Full breakdown below 👇 🧵 1/23
Charles Goddard tweet media
English
104
238
1.6K
509.9K
omer goldman retweetou
Amir Kargaran
Amir Kargaran@amir_nlp·
I really wanted to see the review details. It's clearly above the acceptance threshold of findings for me. When you fall into the cycle of rejection from ARR, it's hard to come out.
Amir Kargaran tweet media
English
0
1
19
2.1K
omer goldman retweetou
AK
AK@_akhaliq·
RefVNLI Towards Scalable Evaluation of Subject-driven Text-to-image Generation
AK tweet media
English
1
52
133
16.9K