February 8, 2026 at 7:00 AM
April 18, 2026 at 3:49 PM
שלום כיתה #נלפ א1
כמה גורמים אפשריים לדעתכם לכישלון המפואר הזה?
#תרגום #בלשנות
#HumanLevelTranslation #neuralempty #NLProc #NLP
כמה גורמים אפשריים לדעתכם לכישלון המפואר הזה?
#תרגום #בלשנות
#HumanLevelTranslation #neuralempty #NLProc #NLP
September 1, 2023 at 5:00 AM
שלום כיתה #נלפ א1
כמה גורמים אפשריים לדעתכם לכישלון המפואר הזה?
#תרגום #בלשנות
#HumanLevelTranslation #neuralempty #NLProc #NLP
כמה גורמים אפשריים לדעתכם לכישלון המפואר הזה?
#תרגום #בלשנות
#HumanLevelTranslation #neuralempty #NLProc #NLP
Headlinese parsing and translation is not a new task (I was involved in such an effort which for reasons not under my control didn’t continue).
Low-ish hanging research project!
#HumanLevelTranslation #neuralempty #NLP
Low-ish hanging research project!
#HumanLevelTranslation #neuralempty #NLP
בשולי, בשפה ההולנדית מקובל לדלג על מילות קישור בכותרות. משום מה, כל מנועי התרגום מתרגמים את הסגנון המעיק לאנגלית, שפה שבה זה *לא עובד*
March 19, 2026 at 6:17 PM
Headlinese parsing and translation is not a new task (I was involved in such an effort which for reasons not under my control didn’t continue).
Low-ish hanging research project!
#HumanLevelTranslation #neuralempty #NLP
Low-ish hanging research project!
#HumanLevelTranslation #neuralempty #NLP
March 18, 2026 at 7:36 PM
September 22, 2026 at 8:20 PM
oh god, I wish #NeuralEmpty was still a thing :( We would have laughed so much with the old twitter clique...
Context: in a alleged microsoft ad "should I get bangs?" got translated to "should I get shagged?" in French (dois-je me faire sauter ?)" instead of "dois-je me faire des franges ?"
Context: in a alleged microsoft ad "should I get bangs?" got translated to "should I get shagged?" in French (dois-je me faire sauter ?)" instead of "dois-je me faire des franges ?"
Le « should I get bangs » (dois-je me faire une frange) qui devient « dois-je me faire sauter » dans une pub Microsoft, j’étais vraiment pas prêt.
February 13, 2025 at 10:54 AM
oh god, I wish #NeuralEmpty was still a thing :( We would have laughed so much with the old twitter clique...
Context: in a alleged microsoft ad "should I get bangs?" got translated to "should I get shagged?" in French (dois-je me faire sauter ?)" instead of "dois-je me faire des franges ?"
Context: in a alleged microsoft ad "should I get bangs?" got translated to "should I get shagged?" in French (dois-je me faire sauter ?)" instead of "dois-je me faire des franges ?"
Back to work... How do we evaluate translation quality in unsegmented video captions?
x.com/alvations/st...
#neuralempty #llm #nlproc #videostreaming #translation #ComputerVision #GenAi #VideoCaptioning
x.com/alvations/st...
#neuralempty #llm #nlproc #videostreaming #translation #ComputerVision #GenAi #VideoCaptioning
#neuralempty #llm #nlproc #machinetranslation | Li Ling Tan
How does one evaluate the quality of translated captions in unsegmented videos?
In typical timestamped subtitle .srt files, we can somehow run translations at segment level and kinda assume independ...
linkedin.com
December 2, 2025 at 4:25 PM
Back to work... How do we evaluate translation quality in unsegmented video captions?
x.com/alvations/st...
#neuralempty #llm #nlproc #videostreaming #translation #ComputerVision #GenAi #VideoCaptioning
x.com/alvations/st...
#neuralempty #llm #nlproc #videostreaming #translation #ComputerVision #GenAi #VideoCaptioning
Findings from #WMT23
Our Chat4 friend is in the winning group across tasks
Most submissions still use from scratch training
Less constrained (low resource) submissions than before
More test suit submissions!
Low resource results TBD (tech issue)
#EMNLP2023 #WMT #neuralEmpty #LLMs
Our Chat4 friend is in the winning group across tasks
Most submissions still use from scratch training
Less constrained (low resource) submissions than before
More test suit submissions!
Low resource results TBD (tech issue)
#EMNLP2023 #WMT #neuralEmpty #LLMs
December 6, 2023 at 1:41 AM
Findings from #WMT23
Our Chat4 friend is in the winning group across tasks
Most submissions still use from scratch training
Less constrained (low resource) submissions than before
More test suit submissions!
Low resource results TBD (tech issue)
#EMNLP2023 #WMT #neuralEmpty #LLMs
Our Chat4 friend is in the winning group across tasks
Most submissions still use from scratch training
Less constrained (low resource) submissions than before
More test suit submissions!
Low resource results TBD (tech issue)
#EMNLP2023 #WMT #neuralEmpty #LLMs
Cross posted on x.com/alvations/st...
Liling Tan on X: "Mark 3 should be stable code now!! #neuralempty #llm #nlproc An adaptation of the model initialization part of @BZhangGo T5Gemma that converts decoder-only models to encoder-decoder @YiTayML UL2 feature for preprocessing data is missing though. https://t.co/Xq8sNjYiJg" / X
Mark 3 should be stable code now!! #neuralempty #llm #nlproc An adaptation of the model initialization part of @BZhangGo T5Gemma that converts decoder-only models to encoder-decoder @YiTayML UL2 feature for preprocessing data is missing though. https://t.co/Xq8sNjYiJg
x.com
January 4, 2026 at 2:31 PM
Cross posted on x.com/alvations/st...
Mark 3 should be stable code now!! #neuralempty #llm #nlproc
An adaptation of the model initialization part of T5Gemma that converts decoder-only models to encoder-decoder
But the UL2 feature for preprocessing data is missing though.
github.com/alvations/AED
An adaptation of the model initialization part of T5Gemma that converts decoder-only models to encoder-decoder
But the UL2 feature for preprocessing data is missing though.
github.com/alvations/AED
January 4, 2026 at 2:30 PM
Mark 3 should be stable code now!! #neuralempty #llm #nlproc
An adaptation of the model initialization part of T5Gemma that converts decoder-only models to encoder-decoder
But the UL2 feature for preprocessing data is missing though.
github.com/alvations/AED
An adaptation of the model initialization part of T5Gemma that converts decoder-only models to encoder-decoder
But the UL2 feature for preprocessing data is missing though.
github.com/alvations/AED
I’m back at the whole #neuralempty eval thingy. Reply with your recent QE or MT eval and I’ll put them on my paper blitz list and I promise to tag you and your co-authors when I post my blitzing posts/summary 😁
What’s the vibe on QE and #LLM as a judge these days?
What’s the vibe on QE and #LLM as a judge these days?
November 18, 2025 at 5:57 PM
I’m back at the whole #neuralempty eval thingy. Reply with your recent QE or MT eval and I’ll put them on my paper blitz list and I promise to tag you and your co-authors when I post my blitzing posts/summary 😁
What’s the vibe on QE and #LLM as a judge these days?
What’s the vibe on QE and #LLM as a judge these days?
