Isaac Yong

A dashboard changed how I make decisions for SPM Chat. Here’s what I track, why trends matter more than vanity metrics, and how it doubles as a control panel.

Most robot controllers have a delay. Commands take time to travel through software stacks, buffers, and hardware. For the Franka Panda arm I was working with in MuJoCo, this delay is exactly 100ms — five control steps at 50Hz. That sounds small. It is not. The problem A standard inverse kinematics (IK) controller does something simple: look at where the target is right now, compute the joint angles that would put the end-effector there, send those angles to the robot. ...

SPM Chat started as a practical question: can an affordable AI system be reliable enough to help Malaysian students revise for SPM? As the project grew, I realized I needed something more rigorous than anecdotal wins and isolated demos. That led to SPM Bench, a benchmark built to measure how well low-cost language models answer SPM multiple-choice questions, and how much a purpose-built retrieval harness actually helps. This write-up summarizes what I learned from the latest benchmark run on this branch, and why the results changed how I think about RAG for education. ...

I built an AI chatbot for SPM students. Here’s what I learned about marketing, small language models, community, and fumbling a beta launch.

MidiBERT-Piano Contributions Compound word(CP) encoding is better than REMI encoding in general BERT-based model outperforms RNN-based model in following downstream tasks: Melody extraction Velocity Prediction Composer Identification Emotion classfication Pretraining perform much better than the model that train from scratch. Future work Implement other pretraining method to further boost the performance and robustness. Personally, I think the recent GLM paper is worth trying.

Hierarchical Perceiver Problems Perception Models are able to process large inputs and largely focused on Global attention. Fourier embeddings must be adjust to fit the modality of data and become memory bottleneck when dealing with high dimensional data Novelties This paper shows that by introducing some degree of locality, it can improve the efficiency of perceiver model. Masked Auto-Encodign(MAE) plays a mojor role in learning positional embeddings Architecture ...

Model architecture for single agent deterministic game which can trained without prior human knowledge about the rules and strategies.. Main Contributions: Monte Carlo Tree Search (MCTS) Solve the exploitation vs exploration dilemma. The use of Representation, Prediction and Dynamic function Prediction functon $f$, predicts policy and value, $p_t$ and $v_t$ Dynamic function $g$, given the current state and action taken, $s_t$ and $a_{t+1}$ predicts the next state and immediate reward, $s_{t+1}$ and $r_{t+1}$ Representation function $h$, convert current state to latent space, $s_t$ Cons: ...

MC-based Competitions MC stands for Multiple Choice. This is a pretty straight forward format because it is just like examinations in school but with slightly advance topics. KMC (Kangaroo Math Competition) IMONST 1 (International Mathematical Olympiad National Selection Test) organized by Committee under Ministry of Education of Malaysia, that is why this test is different to many other competitions organized by other for-profit organizations. aims to select students to represent Malaysia to compete in competitions,, e.g TOT, APMO, and IMO. SASMO (Singapore and Asian Schools Math Olympiad) SMO Education Group organised many competitions and SASMO is one of the most valueable competitions in my opinion as it leads to SIMOC, AIMO, and IJMO. AIMO (Asia International Mathematical Olympiad) One of the common International MO competitions, consists of two rounds, trial and final. Medalists of SASMO will be invited to join AIMO final. 陈景润杯中学数学竞赛 Chen Jing Run’s Cup Secondary School Mathematical Competition organized by Xiamen University Malaysia for secondary school students only and must register through school ASMO (Asian Science and Maths Olympiad) I don’t recommend this competition because of the poor management and rude attitude. Some Comments about ASMO Their past year question sets are pretty terrible too, contain a lot of mistakes and not well-explained at all. Definitely not worth the penny. Proof-based Compeitions This type of competitions is typically harder since and it requires proof writing and takes longer period of time, normally 3 to 4.5 hours. ...

Some of my thoughts when watching MyGeekathon Final 2021. 1. NEVER say no enough time For the sake of your marks, do not say you dont have enough time. All the contestants have the same amount of time and it is sufficient to build a good enough prototype / product. This line just exposed the poor time management of your team to the judges. 2. Understand clearly the judging criteria for the competition Knowing the judging criteria can come in handy when you are deciding what information should be included in the presentation. The presentation time is often short, you want to catch judges’ attention and give them the information they want. If there is a part of the marks allocated for the code, make sure you show some of it (maybe the core parts) in your presentation. ...