MLab
ACMLab is Stanford's premier machine learning club. Its goal is to teach anyone with basic CS experience machine learning. After an intensive ramp-up workshop in the fall, members work on publishing papers at top ML conferences and workshops. We have published 6 workshop papers so far at top conferences and workshops such as ACL and ICLR. Alumni have gone on to Google AI, Stanford ML Group, Stanford NLP Group, and VMWare.
Board
Kenan Hasanaliyev
2025
Co-Director
Conner Takehana
2026
Co-Director
Fall 2023 Schedule
This year, we'll be meeting weekly on Wednesday from 4:30-6:00pm PT.
Sun Oct 08
ACM Info Session
Wed Oct 11
Workshop 1: Shallow Neural Networks
Wed Oct 18
Workshop 2: Deep Neural Networks with Pytorch
Wed Oct 25
Workshop 3: CNNs
Wed Nov 01
Workshop 4: Implementation I
Wed Nov 08
Workshop 5: Implementation II
Wed Nov 15
Workshop 6: Transformers
Wed Nov 29
Onboarding Projects Demo
Recent Projects
SemEVAL
We submitted to the Workshop on Semantic Evaluation's Task 1 (lexical complexity modelling) and Task 8 (automatically extracting measurements from scientific text). Our teams performed competitively on both tasks, including second place in one of the Task 8 subcategories. Our task description papers will appear at SemEval at ACL 2021.
Google BIG-Bench
Members proposed 4 tasks to be used in Google's BIG-Bench challenge. The purpose of this challenge was to create a collaborative benchmark for enourmous language models like GPT-3. MLab submitted tasks about temporal sequences, logic puzzles, sarcasm, and IPA translation.
VQA
We are currently preparing a submission on the ChartQA workshop at CVPR 2021, aiming to automatically parse structured information from diverse chart-based visual representations.