01 · Reason
Reinforcement Learning & Post-Training
I teach language models which earlier steps to focus on, so they reason more accurately with shorter answers.
Machine Learning Engineer & Researcher
Safaeid Hossain Arib
I am a Machine Learning Engineer at ACI PLC, where I build machine learning systems that run in the real world. Before that, I was an AI Engineer at Ontik Technology, developing agentic AI applications.
My latest first-author work, OPASD, teaches reasoning language models where to look. This work is supervised by Dr. Md Mofijul Islam (Amazon GenAI, USA) and Dr. Tasnim Mohiuddin (Qatar Computing Research Institute, Qatar). I am also working with Dr. Ser-Nam Lim (University of Central Florida, USA) on world models.
I earned my B.Sc. in Robotics and Mechatronics Engineering from the University of Dhaka. After graduating, I worked on amodal counting with Dr. Md Mehedi Hasan (University of Dhaka, Bangladesh) at the AVIS Research Lab. As an undergraduate, I worked on sign language translation with Dr. Shafin Rahman (North South University, Bangladesh) and Dr. Sejuti Rahman (New Uzbekistan University, Uzbekistan).
Outside of research, I play basketball, train at the gym, and travel whenever I can. I am always glad to meet curious people, so please feel free to reach out.
I am actively seeking MS/PhD opportunities for Fall 2027.
Research interests
- Language Model Post-Training
- Reinforcement Learning
- Multimodal Learning
- Vision-Language Models
- Embodied AI
- Computer Vision
Research
Most agents learn from outcomes, which tell them whether they succeeded but not why. My work gives an agent richer signals across the three abilities it needs. To reason, it learns where to look, not just which answer is right. To perceive, it fuses modalities so that one view fills in what another misses. To act, it learns why it failed, as a vision–language model turns its mistakes into rewards.
02 · Perceive
Multimodal Learning
I fuse complementary modalities, like video with body pose or time series with text, so models capture what any single view would miss.
03 · Act
Embodied AI
I use vision–language models to explain why a robot failed and turn that feedback into rewards.
News
-
New Our new preprint, Teach Yourself Where to Look, is now on arXiv.
-
Our paper LAGEA: Language Guided Embodied Agents for Robotic Manipulation has been accepted at ICML 2026.
-
Our work T3Time has been accepted at the AAAI 2026 Main Technical Track.
-
SignFormer-GCN has been accepted at the WiML Workshop at NeurIPS 2025.
-
Started as a Machine Learning Engineer at ACI PLC.
-
SignFormer-GCN is officially published in PLOS ONE.
-
Graduated from the Department of Robotics and Mechatronics Engineering, University of Dhaka.
-
Joined Ontik Technology as an AI Engineer.
-
Successfully defended my undergraduate thesis — a huge milestone in my academic journey.
-
Presented our poster Classical machine learning approach for human activity recognition using location data at UbiComp/ISWC 2021.
Publications & Preprints
Google Scholar →
Teach Yourself Where to Look: On-Policy Attention Self-Distillation for Reasoning
On-policy self-distillation tells a student what to predict, but not where to look. OPASD adds solution-conditioned attention distillation, projecting a privileged teacher's attention onto student-visible context. It beats token-only OPSD by 4.98–8.40 points across three Qwen3 sizes while cutting rollout tokens by 73.9%.
LAGEA: Language Guided Embodied Agents for Robotic Manipulation
LAGEA turns structured vision-language-model reflections on a robot's own failures into temporally grounded, progress-based shaping rewards. It localizes the decisive moments of an episode, aligns feedback with visual states, and fades the guidance as the policy improves, outperforming FuRL on Meta-World MT10 and Fetch.
T3Time: Tri-Modal Time Series Forecasting via Adaptive Multi-Head Alignment and Residual Fusion
T3Time reads a time series three ways (temporal, spectral, and through an LLM prompt), gates temporal versus spectral features by forecast horizon, and fuses modalities with adaptively weighted cross-modal attention heads. It sets the best MSE on 7 of 8 long-term benchmarks and stays strong with only 5–10% of the training data.
Counting Through Occlusion: Framework for Open World Amodal Counting
State-of-the-art counters only count what they can see. CountOCC reconstructs the features of hidden objects from visible context and text/exemplar priors, and enforces attention consistency between occluded and unoccluded views. It sets a new state of the art on our occlusion benchmarks FSC-147-OCC and CARPK-OCC, and on CAPTURe-Real.
SignFormer-GCN: Continuous Sign Language Translation Using Spatio-Temporal Graph Convolutional Networks
SignFormer-GCN jointly encodes RGB video with a transformer and skeletal keypoints with an STGCN-LSTM, capturing both broad context and fine-grained body motion. It delivers competitive gloss-free translation on German, American, and Bangla sign language benchmarks with only 9.43M parameters.
Bornil: An Open-Source Sign Language Data Crowdsourcing Platform for AI Enabled Dialect-Agnostic Communication
Bornil is an open-source platform for recording, annotating, and validating multilingual sign language data. It was used to build BornilDB v1.0, the largest Bangladeshi Sign Language dataset (73 hours, 21,000+ samples), with recognition benchmarks for this low-resource setting.
Classical Machine Learning Approach for Human Activity Recognition Using Location Data
Our entry to the Sussex-Huawei Locomotion Challenge 2021 recognizes eight locomotion and transportation activities from smartphone location data using statistical features and a random forest, reaching 78.14% validation accuracy and a 78.28% weighted F1 score.
Experience
Research
-
Jun 2025 – Present
Research Assistant
Dept. of Robotics and Mechatronics Engineering, University of Dhaka · with Dr. Md Mehedi Hasan and Md. Jubair Ahmed Sourov
- Designed and implemented CountOCC, an open-world amodal counting framework, and built occlusion-augmented evaluation benchmarks for structured and unstructured scenes.
- Co-authored the resulting WACV 2026 paper.
-
Jan 2023 – Dec 2023
Undergraduate Research Assistant
Dept. of Robotics and Mechatronics Engineering, University of Dhaka · with Dr. Shafin Rahman and Dr. Sejuti Rahman
- Designed and implemented a novel architecture for continuous sign language translation across Bangla, American, and German benchmarks.
- Co-authored the resulting work, SignFormer-GCN, published in PLOS ONE.
Industry
-
Sep 2025 – Present
Machine Learning Engineer
ACI PLC · Dhaka, Bangladesh
- Developed a few-shot classification pipeline for fine-grained recognition of visually similar products.
- Building a lightweight text-to-speech model for few-shot voice cloning and low-resource language adaptation.
- Improving a video-based truck identification pipeline by enhancing low-resolution license plate imagery for more reliable detection and OCR.
- Developing an agentic assistant for meeting scheduling, document summarization, and conversational search across company files through Drive and Calendar integration.
-
Feb 2024 – Aug 2025
AI Engineer
Ontik Technology · Dhaka, Bangladesh
- Developed agentic B2B automation systems that generated social-media insights and automated client-ready reporting, substantially reducing manual workflows.
- Built agentic ETL pipelines for COBOL copybook parsing and legacy modernization, enabling AI-driven analytics and supporting demonstrations to consulting firms including Oliver Wyman.
- Developed a patient-facing OCR tool for prescription scanning, medication scheduling, and alternative-medicine lookup under Dr. Javed Mostafa.
Teaching
-
Jun 2024 – Aug 2024
Instructor
National Camp, Bangladesh Artificial Intelligence Olympiad
- Designed and delivered lessons on classification algorithms and evaluation metrics for high school students, and created quizzes to assess their understanding.
