Dylan Dahlquist

Scientist. Previous: @HMGaming, @RedBull, @NH_RC, @CSIPacific. (opinions/tweets are my own)

[Antonis Elia @Antoniselia1992]

Our latest study is now published in @JPhysiol

In this paper we examined the skeletal muscle characteristics of competitive breath-hold divers 🤿

A true international team effort 🇸🇪🇨🇦🏴󠁧󠁢󠁥󠁮󠁧󠁿

You can access the paper here 👇

https://t.co/FvyECV8D2q

#freediving #apnea #physiology

[Alexandria Ocasio-Cortez @AOC]

@NancyMace The only fraudulent activity here is your covering up how you were endangering survivor info, attacking one of the few other survivors in Congress, & refusing to fix serious flaws that endanger victims without consent.

Fix it and people will vote YES. https://t.co/zBhiwLjvE2

[Specialized Training™️ @Tatopek1]

🆕"HRV is most often used independently, without complementary data (e.g., ACC + blood pressure), & its interpretation is usually limited to a vague ‘sympathetic-vagal balance' "
👉N. Bourdillon @GregoireMillet1 et al, 2026 🇨🇭
📂@FrontSportsAL Open Access: https://t.co/4aFsc5Zkv6

[Yan Holtz @R_Graph_Gallery]

😀🍾 Big, big news!! 🎉😱

➡️ I'm starting a new #dataviz project with @CedScherer

It's an interactive online course designed to turn you into a ggplot2 dataviz expert!

➡️ https://t.co/HxJb2xar9Z

Pre-launch until Sunday. If you like making charts with R, you'll love it!

[Tadej Debevec @DebevecTadej]

New in @MSSEonline

Is the Classification of Athletes Based on Force-Velocity Profiling from Countermovement Jumps Influenced by the Choice of Method?

https://t.co/QHeiURBn2t

[Dylan Dahlquist @dyldahl]

@rmarti33 I need it.

[Dylan Dahlquist @dyldahl]

@AndrewHires @Skrillex Tear lost drop 🔥

[Mackenzie Weygandt Mathis, PhD @TrackingActions]

✨ Introducing a new #SOTA action recognition large multimodal language model: #LLaVAction!

Understanding human behavior requires recognizing actions—a challenging task given the complexity of behavior. Large multimodal language models (#MLLMs) offer a promising path forward, but how well do they perform in action recognition?

In our latest work - by @shaokaiyeah, Haozhe Qi, @TrackingPlumes and me | @EPFL_en - we rigorously evaluate and enhance MLLMs for action recognition in a real-world and challenging settings- egocentric views in the kitchen! 🧑‍🍳🔪🧽🤖

👀 We find that developing a multi-question-answer (#MQA) task serves as a valuable intermediate step in training (and evaluating) MLLMs for action understanding. Namely, we introduce EPIC-KITCHENS-100-MQA, a reformulation of the highly challenging EPIC-KITCHENS-100 dataset into a video multiple-choice question-answering task which allows for rigorous benchmarking of MLLMs in this task.

Next, we propose methods that substantially improve MLLM performance, and even achieving state-of-the-art results 🏆 (#SOTA) on the EPIC-KITCHENS-100 validation set 🔥✨.

Our approach also outperforms GPT-4o by 21 points in accuracy on EPIC-KITCHENS-100-MQA and demonstrates improvements across other action-related video benchmarks, including #VideoMME, #PerceptionTest, and #MVBench.

Our #LLaVAction-7B and -0.5B models can do #MQA and, critically, can do video captioning! 🙏🚀

As MLLMs become central to AI-driven video understanding in such real-world settings, ensuring their robustness in real-world tasks is critical. Excited to push the boundaries of multimodal AI further! 💪

🇨🇭We could not have done this without the amazing support of #SwissAI: the Swiss AI Initiative & the Swiss National Supercomputing Centre (#CSCS). @EPFL_AI_Center

#ProjectPage: https://t.co/BlmhrYKHXH
📝 #arXivPaper: https://t.co/yMOdYYw2Yv
👩‍💻💻 GitHub code & Google #ColabDemo: https://t.co/eO3LO7SOiJ
🤗 Hugging Face models (use with transformers): https://t.co/1V7Z83WE33

#AI #MultimodalLearning #ActionRecognition #EPICKITCHENS100 #MLLM #LLaVAction #VideoCaptioning #VLMs

[HEaT Lab @kiwiheatlab]

New in @NatureComms from @robertdmeade and all-star cast 🤩:

Meta-analysis of heat-induced changes in cardiac function from over 400 laboratory-based heat exposure studies @HEPRU_uOttawa 👇

https://t.co/cNlB2V1seV

[Yan Holtz @R_Graph_Gallery]

🔥 New Python library for #dataviz!

"morethemes" gives your graph a better theme in 1 line of code!

It currently offers 10 themes inspired by famous journals such as The Wall Street Journal, andThe Economist.

Learn more: https://t.co/4D4muqdkPw

Congrats @joseph_barbier !

[Brett Singer MS,RD,CSSD @BSinger10]

Can creatine enhance post-exercise muscle glycogen synthesis?

Article: Creatine ingestion augments dietary carbohydrate mediated muscle glycogen supercompensation during the initial 24 h of recovery following prolonged exhaustive exercise in humans

Link: https://t.co/CA0j5LgVjg

[Nicholas Fabiano, MD @NTFabiano]

Our paper was just published in European Neuropsychopharmacology.

🧵1/10

[DeepLabCut 🦄 @DeepLabCut]

👀 We hit a major milestone: the PR for DLC3 is merged into main, and rc7 is up for final testing: https://t.co/ISfD5rnOi5

- this is the most powerful, most new-user (+ new GUI features!) & ML expert friendly code, & the highest performance ... ready to #AnimalPose?! 🚀

[Georgi Gerganov @ggerganov]

Today I was sent the following cool demo:

Two AI agents on a phone call realize they’re both AI and switch to a superior audio signal ggwave

[Tadej Debevec @DebevecTadej]

Effect of condensed heat acclimation on thermophysiological adaptations, hypoxic cross-tolerance, exercise performance, and deacclimation | Journal of Applied Physiology | ⁦@japplphysiol⁩ https://t.co/eTDwrhNilu

[Dylan Dahlquist @dyldahl]

The birth of exercise physiology - would recommend this piece to all current and aspiring ex physiologist, sport scientists, etc.

#physiology #exercise #sportscience #research https://t.co/37rAlkxpK2

[Ash Willmott @AshWillmott]

Still open:

📢3-year (>£19k year) PhD at @ARU_Sport

🏹Topic: Establishing a sport-specific, evidence-based #archery classification system for athletes with a vision impairment

🔗Link: https://t.co/VUDbPSY2s6…

📅Deadline: 16/3/25

✉️ informal chat/Qs - [email protected]

[Tadej Debevec @DebevecTadej]

Conférence Internationale : "Using Environmental Stressors to Achieve Peak Performance: From Evidence to Avenues to Explore" | INSEP https://t.co/rmSVR94cum

[Jem Arnold @jem_arnold]

In our recent meta-analysis with trained athletes, did the athletes who improved their VO2max more also improve their endurance performance more?🤔

No! 😮
In this dataset, +1% change in VO2max was associated with only +0.04% change in TT performance. But why? 1/15🧵

[Tadej Debevec @DebevecTadej]

A Practical Deep Learning Model for Core Temperature Prediction of Specialized Workers in High-Temperature Environments

https://t.co/RVKWz5KhYw