11 papers
ConRetroBert: EMA Stabilized Dual Encoders for Template-Based Single-Step Retrosynthesis
Mohammad Jahid Ibna Basher, Ali Khodabandeh Yalabadi, Ivan Garibay +1
Template based single step retrosynthesis predicts reactants by selecting and applying an explicit reaction template, making each prediction traceable to a chemical transformation…
When Policies Cannot Be Retrained: A Unified Closed-Form View of Post-Training Steering in Offline Reinforcement Learning
Elias Hossain, Mohammad Jahid Ibna Basher, Ivan Garibay +2
Offline reinforcement learning (RL) can learn effective policies from fixed datasets, but deployment objectives may change after training, and in many applications the trained acto…
LiME: Lightweight Mixture of Experts for Efficient Multimodal Multi-task Learning
Md Kowsher, Haris Mansoor, Nusrat Jahan Prottasha +4
MoE-PEFT methods combine Mixture of Experts with parameter-efficient fine-tuning for multi-task adaptation, but require separate adapters per expert causing trainable parameters to…
Monkey Jump : MoE-Style PEFT for Efficient Multi-Task Learning
Nusrat Jahan Prottasha, Md Kowsher, Chun-Nam Yu +2
Mixture-of-experts variants of parameter-efficient fine-tuning enable per-token specialization, but they introduce additional trainable routers and expert parameters, increasing me…
FlowNIB: An Information Bottleneck Analysis of Bidirectional vs. Unidirectional Language Models
Md Kowsher, Nusrat Jahan Prottasha, Shiyun Xu +4
Bidirectional language models have better context understanding and perform better than unidirectional models on natural language understanding tasks, yet the theoretical reasons b…
LLM-Mixer: Multiscale Mixing in LLMs for Time Series Forecasting
Md Kowsher, Md. Shohanur Islam Sobuj, Nusrat Jahan Prottasha +3
Time series forecasting remains a challenging task, particularly in the context of complex multiscale temporal patterns. This study presents LLM-Mixer, a framework that improves fo…