Katya Artemova
@katya-art.bsky.social
#NLProc researcher @ Toloka AI, ex LMU, ex HSE
Low resource languages | culture-aware LLMs | machine-generated test detection
Low resource languages | culture-aware LLMs | machine-generated test detection
Our paper "𝗕𝗲𝗲𝗺𝗼: 𝗕𝗲𝗻𝗰𝗵𝗺𝗮𝗿𝗸 𝗼𝗳 𝗘𝘅𝗽𝗲𝗿𝘁-𝗲𝗱𝗶𝘁𝗲𝗱 𝗠𝗮𝗰𝗵𝗶𝗻𝗲-𝗴𝗲𝗻𝗲𝗿𝗮𝘁𝗲𝗱 𝗢𝘂𝘁𝗽𝘂𝘁𝘀" will be presented at #NAACL2025!
Beemo is one of the first benchmarks designed to evaluate AI detector performance on mixed-authorship content — texts generated by LLMs and edited by human experts or refined by other models.
Beemo is one of the first benchmarks designed to evaluate AI detector performance on mixed-authorship content — texts generated by LLMs and edited by human experts or refined by other models.
April 29, 2025 at 7:49 PM
Our paper "𝗕𝗲𝗲𝗺𝗼: 𝗕𝗲𝗻𝗰𝗵𝗺𝗮𝗿𝗸 𝗼𝗳 𝗘𝘅𝗽𝗲𝗿𝘁-𝗲𝗱𝗶𝘁𝗲𝗱 𝗠𝗮𝗰𝗵𝗶𝗻𝗲-𝗴𝗲𝗻𝗲𝗿𝗮𝘁𝗲𝗱 𝗢𝘂𝘁𝗽𝘂𝘁𝘀" will be presented at #NAACL2025!
Beemo is one of the first benchmarks designed to evaluate AI detector performance on mixed-authorship content — texts generated by LLMs and edited by human experts or refined by other models.
Beemo is one of the first benchmarks designed to evaluate AI detector performance on mixed-authorship content — texts generated by LLMs and edited by human experts or refined by other models.
Join our Multilingual Models webinar on April 15. We'll explore current research on multilingual AI systems with a panel of industry experts. Learn more and register here: bit.ly/4hKHiaE
Multilingual Models: Development and Evaluation
What can we learn from current research on multilingual models? Join us for an in-depth webinar exploring the latest advancements in multilingual AI systems.
bit.ly
March 27, 2025 at 9:48 AM
Join our Multilingual Models webinar on April 15. We'll explore current research on multilingual AI systems with a panel of industry experts. Learn more and register here: bit.ly/4hKHiaE
Preference prediction & explanation at ELOQUENT, #CLEF2025
We invite systems that can predict human preferences for different LLM outputs and explain their predictions across five criteria: relevance, naturalness, truthfulness, safety, and overall quality.
Dev Stage: Feb 3 – Mar 2
We invite systems that can predict human preferences for different LLM outputs and explain their predictions across five criteria: relevance, naturalness, truthfulness, safety, and overall quality.
Dev Stage: Feb 3 – Mar 2
February 11, 2025 at 7:44 PM
Preference prediction & explanation at ELOQUENT, #CLEF2025
We invite systems that can predict human preferences for different LLM outputs and explain their predictions across five criteria: relevance, naturalness, truthfulness, safety, and overall quality.
Dev Stage: Feb 3 – Mar 2
We invite systems that can predict human preferences for different LLM outputs and explain their predictions across five criteria: relevance, naturalness, truthfulness, safety, and overall quality.
Dev Stage: Feb 3 – Mar 2
Attending #COLING2025 in Abu Dhabi? Join us for a Hands-on Tutorial: Labeling With LLMs and Human-in-the-Loop.
You’ll learn how to speed up data annotation and reduce costs and human workload.
Discover more: toloka.ai/events/tolok...
You’ll learn how to speed up data annotation and reduce costs and human workload.
Discover more: toloka.ai/events/tolok...
Toloka
Powering AI with human insight
toloka.ai
January 7, 2025 at 1:09 PM
Attending #COLING2025 in Abu Dhabi? Join us for a Hands-on Tutorial: Labeling With LLMs and Human-in-the-Loop.
You’ll learn how to speed up data annotation and reduce costs and human workload.
Discover more: toloka.ai/events/tolok...
You’ll learn how to speed up data annotation and reduce costs and human workload.
Discover more: toloka.ai/events/tolok...
Reposted by Katya Artemova
Analysis finds over 54% of longer English-language posts on LinkedIn are likely AI-generated; LinkedIn says it doesn't track how many posts are created by AI (Kate Knibbs/Wired)
Main Link | Techmeme Permalink
Main Link | Techmeme Permalink
November 26, 2024 at 11:06 PM
Analysis finds over 54% of longer English-language posts on LinkedIn are likely AI-generated; LinkedIn says it doesn't track how many posts are created by AI (Kate Knibbs/Wired)
Main Link | Techmeme Permalink
Main Link | Techmeme Permalink