Ọládípọ̀ AF

Guided. Gifted. I will ask for your support someday. I hope I will have earned it.

Research Scientist @ServiceNow

[Ọládípọ̀ AF @theyorubayesian]

Many little things point towards Nigeria falling apart. It bothers me.

https://t.co/UOpUA0o0Vi

[The Nigerian Thinker @TheNGThinker]

𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗮𝗿𝗲 𝗻𝗼𝘄 𝗼𝗽𝗲𝗻 𝗳𝗼𝗿 𝗧𝗵𝗲 𝗡𝗶𝗴𝗲𝗿𝗶𝗮𝗻 𝗧𝗵𝗶𝗻𝗸𝗲𝗿 𝗖𝗮𝗺𝗽𝘂𝘀 𝗔𝗺𝗯𝗮𝘀𝘀𝗮𝗱𝗼𝗿 𝗣𝗿𝗼𝗴𝗿𝗮𝗺.

We're looking for thoughtful students who care about ideas, conversations, leadership, and the future of Nigeria.

As a The Nigeria Thinker Campus Ambassador, you'll help connect conversations, represent your campus, and contribute to a growing network of students engaging with important questions and ideas.

𝘛𝘩𝘪𝘯𝘬 𝘺𝘰𝘶'𝘳𝘦 𝘵𝘩𝘦 𝘳𝘪𝘨𝘩𝘵 𝘧𝘪𝘵?

𝗔𝗽𝗽𝗹𝘆 𝗻𝗼𝘄: https://t.co/cKsMQvUHGt

#TheNigerianThinker

[The Nigerian Thinker @TheNGThinker]

𝗜𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗶𝗻𝗴 𝗧𝗵𝗲 𝗡𝗶���𝗲𝗿𝗶𝗮𝗻 𝗧𝗵𝗶𝗻𝗸𝗲𝗿

Nigeria does not have a shortage of opinions. It has a shortage of rigorous thinking and of institutions willing to sustain it.

The Nigerian Thinker exists to cultivate thoughtful conversations, challenge assumptions, and shape the quality of thinking that influences the future of our nation.

Because the Nigeria we build tomorrow depends on the quality of thought we encourage today.

#TheNigerianThinker

[Ọládípọ̀ AF @theyorubayesian]

knee slide for first goal. proper chels.

[Conn @ConnCFC]

ESTEVAO FIRST CHELSEA GOAL IS THE WINNER VS LIVERPOOL

[Nouha Dziri @nouhadziri]

🤯 We noticed that many failures stem not from lack of knowledge but from overthinking. Models often find the right answer early in CoT, but spiral into self-corrections and abandon correct solutions. This challenges the assumption:

More CoT ≠ better results

Sometimes the models' self-correction mechanisms can inadvertently backfire

[Ayoyemi Oladipo @Ayoyemi_OK]

@richardokunolar Not a newspaper yet, but I publish good news every Friday on my LinkedIn page:

https://t.co/TWJ7AbTI6h

[Ọládípọ̀ AF @theyorubayesian]

@toluclassics 😅

[Ọládípọ̀ AF @theyorubayesian]

@JessicaOjo__ @Mila_Quebec Congratulations! ✨

[Xiang Yue @xiangyue96]

🚀 Introducing MMMU-Pro: A more robust version of MMMU
https://t.co/zAVJrPO8CV
After launching MMMU, we received valuable feedback from the community:

1️⃣ Some questions were answerable without even seeing the images.
2️⃣ Models didn’t always "know" the answer but found shortcuts from the options provided.
3️⃣ Performance was heavily tied to LLMs, with minimal impact from the vision module.

To tackle these issues, we implemented the following improvements:

🔍 1. Filtering Text-Only Answerable Questions
🔄 2. Augmenting Candidate Options up to 10 by Human Experts.
🖼️ 3. Vision-Only Input Setting: where questions are embedded directly in images, requiring the model to rely purely on visual input.

✨ Why We Added Vision-Only Input Setting?

1. From a foundational perspective, this setting forces AI to genuinely "see" and "read" at the same time—challenging a core human cognitive skill: the seamless integration of visual and textual information.

2. From an application standpoint, this approach mirrors how users naturally interact with AI systems—by sharing screenshots or photos, without meticulously separating text from images.

📊 Key Results:

Performance on MMMU-Pro is notably lower compared to MMMU, ranging from 16.8% to 26.9% across various models. The ranking of models is generally similar to the original but we also observe less robust ones— for example, GPT-4o mini proved less robust than GPT-4o and other proprietary models, showing significant drops in performance on the augmented set.

🔬 More in-depth analysis can be found in the threads below! 👇

[Ọládípọ̀ AF @theyorubayesian]

Hi @FuadXIV. I'm trying to reach out to you. I sent a DM.

[Ọládípọ̀ AF @theyorubayesian]

“Sometimes it falls upon a generation to be great. You can be that great generation. Let your greatness blossom. Of course, the task will not be easy. But not to do this would be a crime against humanity, against which I ask all humanity now to rise up.” — Nelson Mandela

[Ọládípọ̀ AF @theyorubayesian]

Once again:

https://t.co/WY6KUUkI9n

[Ndi Kato (Ungwan Masara's finest) @YarKafanchan]

It is not abysmal. Nigeria simply doesn't have the population it claims to have.

[Ọládípọ̀ AF @theyorubayesian]

Grow with me.

https://t.co/R0C4F6yMfT

[Ọládípọ̀ AF @theyorubayesian]

@Kowalski93 #Hope23 as a nod to Abiola's #Hope93.

[Ọládípọ̀ AF @theyorubayesian]

A big part of winning elections in Nigeria is convincing people you can win the election.

[Ọládípọ̀ AF @theyorubayesian]

@aladelekan187 @temilaj You didn't perform any validation while training?

[Ọládípọ̀ AF @theyorubayesian]

@temilaj Knowing how to spend those waiting hours is important. I've been training across 2 GPUs for 4 days now, tracking learning rate, loss and ppl in Neptune.

[Ọládípọ̀ AF @theyorubayesian]

@emeka_boris https://t.co/eD7tPgF2It

[Ọládípọ̀ AF @theyorubayesian]

@emeka_boris https://t.co/x8r2IpTGpV

[Ọládípọ̀ AF @theyorubayesian]

@bleso_a @emeka_boris @AlaniPoe https://t.co/eD7tPgF2It

[Masakhane @MasakhaneNLP]

Hello! happy to announce the Masakhane study group currently led by @Oladipo_AF and @Shmuhammadd where we are delving into CMU’s Neural Nets for NLP 2021 course. You don’t have to be an expert 😎, just be committed to learn!