Md Kowsher

Nusrat Jahan Prottasha, Saydul Akbar Murad, Abu Jafar Md Muzahid et al.

Machine learning is the study of computer algorithms that can automatically improve based on data and experience. Machine learning algorithms build a model from sample data, called training data, to make predictions or judgments without being explicitly programmed to do so. A variety of wellknown machine learning algorithms have been developed for use in the field of computer science to analyze data. This paper introduced a new machine learning algorithm called impact learning. Impact learning is a supervised learning algorithm that can be consolidated in both classification and regression problems. It can furthermore manifest its superiority in analyzing competitive data. This algorithm is remarkable for learning from the competitive situation and the competition comes from the effects of autonomous features. It is prepared by the impacts of the highlights from the intrinsic rate of natural increase (RNI). We, moreover, manifest the prevalence of the impact learning over the conventional machine learning algorithm.

Md Kowsher, Nusrat Jahan Prottasha, Prakash Bhat

The rapid advancements in Large Language Models (LLMs) have revolutionized natural language processing (NLP) and related fields. However, fine-tuning these models for specific tasks remains computationally expensive and risks degrading pre-learned features. To address these challenges, we propose Propulsion, a novel parameter efficient fine-tuning (PEFT) method designed to optimize task-specific performance while drastically reducing computational overhead. Inspired by the concept of controlled adjustments in physical motion, Propulsion selectively re-scales specific dimensions of a pre-trained model, guiding output predictions toward task objectives without modifying the model's parameters. By introducing lightweight, trainable Propulsion parameters at the pre-trained layer, we minimize the number of parameters updated during fine-tuning, preventing overfitting or overwriting of existing knowledge. Our theoretical analysis, supported by Neural Tangent Kernel (NTK) theory, shows that Propulsion approximates the performance of full fine-tuning with far fewer trainable parameters. Empirically, Propulsion reduces the parameter count from 355.3 million to just 0.086 million, achieving over a 10x reduction compared to standard approaches like LoRA while maintaining competitive performance across benchmarks.

Nusrat Jahan Prottasha, Md Kowsher, Chun-Nam Yu et al.

Mixture-of-experts variants of parameter-efficient fine-tuning enable per-token specialization, but they introduce additional trainable routers and expert parameters, increasing memory usage and training cost. This undermines the core goal of parameter-efficient fine-tuning. We propose Monkey Jump, a method that brings mixture-of-experts-style specialization to parameter-efficient fine-tuning without introducing extra trainable parameters for experts or routers. Instead of adding new adapters as experts, Monkey Jump treats the adapters already present in each Transformer block (such as query, key, value, up, and down projections) as implicit experts and routes tokens among them. Routing is performed using k-means clustering with exponentially moving averaged cluster centers, requiring no gradients and no learned parameters. We theoretically show that token-wise routing increases expressivity and can outperform shared adapters by avoiding cancellation effects. Across multi-task experiments covering 14 text, 14 image, and 19 video benchmarks, Monkey Jump achieves competitive performance with mixture-of-experts-based parameter-efficient fine-tuning methods while using 7 to 29 times fewer trainable parameters, up to 48 percent lower memory consumption, and 1.5 to 2 times faster training. Monkey Jump is architecture-agnostic and can be applied to any adapter-based parameter-efficient fine-tuning method.

Md Kowsher, Haris Mansoor, Nusrat Jahan Prottasha et al.

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 scale linearly with expert count and limiting applicability to adapter-based architectures. We propose LiME (Lightweight Mixture of Experts), which achieves expert specialization through lightweight modulation rather than adapter replication. Instead of separate adapters, LiME uses a single shared PEFT module and modulates its output with lightweight expert vectors, reducing expert parameters while generalizing to any PEFT method. Notably, LiME introduces zero-parameter routing by leveraging existing frozen and adapted representations eliminating learned router parameters typically required per layer. Theoretically, we prove that (i) more experts preserve more task-relevant information and (ii) modulation approximates full expert-specific PEFT with bounded error. LiME further incorporates n-gram windowed routing and adaptive expert selection (Auto Top-K) based on routing confidence. Experiments on MMT-47, a multimodal multi-task benchmark with 47 tasks spanning text, image, and video, demonstrate that LiME achieves competitive or superior performance while using up to 4x fewer trainable parameters and up to 29% faster training compared to corresponding MoE-PEFT baselines.

Shahriar Kabir Nahin, Rabindra Nath Nandi, Sagor Sarker et al.

In this paper, we present TituLLMs, the first large pretrained Bangla LLMs, available in 1b and 3b parameter sizes. Due to computational constraints during both training and inference, we focused on smaller models. To train TituLLMs, we collected a pretraining dataset of approximately ~37 billion tokens. We extended the Llama-3.2 tokenizer to incorporate language- and culture-specific knowledge, which also enables faster training and inference. There was a lack of benchmarking datasets to benchmark LLMs for Bangla. To address this gap, we developed five benchmarking datasets. We benchmarked various LLMs, including TituLLMs, and demonstrated that TituLLMs outperforms its initial multilingual versions. However, this is not always the case, highlighting the complexities of language adaptation. Our work lays the groundwork for adapting existing multilingual open models to other low-resource languages. To facilitate broader adoption and further research, we have made the TituLLMs models and benchmarking datasets publicly available (https://huggingface.co/collections/hishab/titulm-llama-family-6718d31fc1b83529276f490a).

Nusrat Jahan Prottasha, Md Kowsher, Hafijur Raman et al.

User profile modeling plays a key role in personalized systems, as it requires building accurate profiles and updating them with new information. In this paper, we present two high-quality open-source user profile datasets: one for profile construction and another for profile updating. These datasets offer a strong basis for evaluating user profile modeling techniques in dynamic settings. We also show a methodology that uses large language models (LLMs) to tackle both profile construction and updating. Our method uses a probabilistic framework to predict user profiles from input text, allowing for precise and context-aware profile generation. Our experiments demonstrate that models like Mistral-7b and Llama2-7b perform strongly in both tasks. LLMs improve the precision and recall of the generated profiles, and high evaluation scores confirm the effectiveness of our approach.

Md Kowsher, Weiwei Zhan, Chen Chen

Landslide detection from high resolution satellite imagery is a critical task for disaster response and risk assessment, yet the relative effectiveness of modern segmentation architectures and finetuning strategies for this problem remains insufficiently understood. In this work, we present a systematic benchmarking study of convolutional neural networks, transformer based segmentation models, and large pre-trained foundation models for landslide detection. Using the Globally Distributed Coseismic Landslide Dataset (GDCLD) dataset, we evaluate representative CNN- and transformer-based segmentation models alongside large pretrained foundation models under consistent training and evaluation protocols. In addition, we compare full fine-tuning with parameter-efficient fine-tuning methods, including LoRA and AdaLoRA, to assess their performance efficiency tradeoffs. Experimental results show that transformer-based models achieve strong segmentation performance, while parameter efficient finetuning reduces trainable parameters by up to 95% with comparable accuracy to full finetuning. We further analyze generalization under distribution shift by comparing validation and held-out test performance.

Nusrat Jahan Prottasha, Asif Mahmud, Md. Shohanur Islam Sobuj et al.

Large Language Models (LLMs) are gaining significant popularity in recent years for specialized tasks using prompts due to their low computational cost. Standard methods like prefix tuning utilize special, modifiable tokens that lack semantic meaning and require extensive training for best performance, often falling short. In this context, we propose a novel method called Semantic Knowledge Tuning (SK-Tuning) for prompt and prefix tuning that employs meaningful words instead of random tokens. This method involves using a fixed LLM to understand and process the semantic content of the prompt through zero-shot capabilities. Following this, it integrates the processed prompt with the input text to improve the model's performance on particular tasks. Our experimental results show that SK-Tuning exhibits faster training times, fewer parameters, and superior performance on tasks such as text classification and understanding compared to other tuning methods. This approach offers a promising method for optimizing the efficiency and effectiveness of LLMs in processing language tasks.

Nusrat Jahan Prottasha, Upama Roy Chowdhury, Shetu Mohanto et al.

Large models such as Large Language Models (LLMs) and Vision Language Models (VLMs) have transformed artificial intelligence, powering applications in natural language processing, computer vision, and multimodal learning. However, fully fine-tuning these models remains expensive, requiring extensive computational resources, memory, and task-specific data. Parameter-Efficient Fine-Tuning (PEFT) has emerged as a promising solution that allows adapting large models to downstream tasks by updating only a small portion of parameters. This survey presents a comprehensive overview of PEFT techniques, focusing on their motivations, design principles, and effectiveness. We begin by analyzing the resource and accessibility challenges posed by traditional fine-tuning and highlight key issues, such as overfitting, catastrophic forgetting, and parameter inefficiency. We then introduce a structured taxonomy of PEFT methods -- grouped into additive, selective, reparameterized, hybrid, and unified frameworks -- and systematically compare their mechanisms and trade-offs. Beyond taxonomy, we explore the impact of PEFT across diverse domains, including language, vision, and generative modeling, showing how these techniques offer strong performance with lower resource costs. We also discuss important open challenges in scalability, interpretability, and robustness, and suggest future directions such as federated learning, domain adaptation, and theoretical grounding. Our goal is to provide a unified understanding of PEFT and its growing role in enabling practical, efficient, and sustainable use of large models.

Md Kowsher, Tara Esmaeilbeig, Chun-Nam Yu et al.

We propose RoCoFT, a parameter-efficient fine-tuning method for large-scale language models (LMs) based on updating only a few rows and columns of the weight matrices in transformers. Through extensive experiments with medium-size LMs like BERT and RoBERTa, and larger LMs like Bloom-7B, Llama2-7B, and Llama2-13B, we show that our method gives comparable or better accuracies than state-of-art PEFT methods while also being more memory and computation-efficient. We also study the reason behind the effectiveness of our method with tools from neural tangent kernel theory. We empirically demonstrate that our kernel, constructed using a restricted set of row and column parameters, are numerically close to the full-parameter kernel and gives comparable classification performance. Ablation studies are conducted to investigate the impact of different algorithmic choices, including the selection strategy for rows and columns as well as the optimal rank for effective implementation of our method.

Md Kowsher, Nusrat Jahan Prottasha, Prakash Bhat et al.

This paper argues that generating output tokens is more effective than using pooled representations for prediction tasks because token-level generation retains more mutual information. Since LLMs are trained on massive text corpora using next-token prediction, generation aligns naturally with their learned behavior. Using the Data Processing Inequality (DPI), we provide both theoretical and empirical evidence supporting this claim. However, autoregressive models face two key challenges when used for prediction: (1) exposure bias, where the model sees ground truth tokens during training but relies on its own predictions during inference, leading to errors, and (2) format mismatch, where discrete tokens do not always align with the tasks required output structure. To address these challenges, we introduce PredGen(Predicting Through Generating), an end to end framework that (i) uses scheduled sampling to reduce exposure bias, and (ii) introduces a task adapter to convert the generated tokens into structured outputs. Additionally, we introduce Writer-Director Alignment Loss (WDAL), which ensures consistency between token generation and final task predictions, improving both text coherence and numerical accuracy. We evaluate PredGen on multiple classification and regression benchmarks. Our results show that PredGen consistently outperforms standard baselines, demonstrating its effectiveness in structured prediction tasks.

Md Kowsher, Nusrat Jahan Prottasha, Chun-Nam Yu et al.

The success of self-attention lies in its ability to capture long-range dependencies and enhance context understanding, but it is limited by its computational complexity and challenges in handling sequential data with inherent directionality. This work introduces a shared weight self-attention-based BERT model that only learns one weight matrix for (Key, Value, and Query) representations instead of three individual matrices for each of them. Our shared weight attention reduces the training parameter size by more than half and training time by around one-tenth. Furthermore, we demonstrate higher prediction accuracy on small tasks of GLUE over the BERT baseline and in particular a generalization power on noisy and out-of-domain data. Experimental results indicate that our shared self-attention method achieves a parameter size reduction of 66.53% in the attention block. In the GLUE dataset, the shared weight self-attention-based BERT model demonstrates accuracy improvements of 0.38%, 5.81%, and 1.06% over the standard, symmetric, and pairwise attention-based BERT models, respectively. The model and source code are available at Anonymous.

Md Kowsher, Md. Shohanur Islam Sobuj, Nusrat Jahan Prottasha et al.

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 forecasting accuracy through the combination of multiscale time-series decomposition with pre-trained LLMs (Large Language Models). LLM-Mixer captures both short-term fluctuations and long-term trends by decomposing the data into multiple temporal resolutions and processing them with a frozen LLM, guided by a textual prompt specifically designed for time-series data. Extensive experiments conducted on multivariate and univariate datasets demonstrate that LLM-Mixer achieves competitive performance, outperforming recent state-of-the-art models across various forecasting horizons. This work highlights the potential of combining multiscale analysis and LLMs for effective and scalable time-series forecasting.

Md Kowsher, Ali O. Polat, Ehsan Mohammady Ardehaly et al.

This paper presents a theoretical framework explaining why fine tuning small, randomly selected subnetworks (slices) within pre trained models can be sufficient for downstream adaptation. We prove that pretrained networks exhibit a universal winning slice property arising from two phenomena: (1) spectral balance the eigenspectra of different weight matrix slices are remarkably similar; and (2) high task energy their backbone representations retain rich, task relevant features. This leads to the Universal Winning Slice Hypothesis, which provides a theoretical foundation for parameter efficient fine tuning (PEFT) in large scale models. Inspired by this, we propose SliceFine, a PEFT method that exploits this inherent redundancy by updating only selected slices of the original weights introducing zero new parameters, unlike adapter-based approaches. Empirically, SliceFine matches the performance of state of the art PEFT methods across language and vision tasks, while significantly improving training speed, memory efficiency, and model compactness. Our work bridges theory and practice, offering a theoretically grounded alternative to existing PEFT techniques.

Sina Abdidizaji, Md Kowsher, Niloofar Yousefi et al.

In the era of digitalization, as individuals increasingly rely on digital platforms for communication and news consumption, various actors employ linguistic strategies to influence public perception. While models have become proficient at detecting explicit patterns, which typically appear in texts as single remarks referred to as utterances, such as social media posts, malicious actors have shifted toward utilizing implicit influential verbal patterns embedded within conversations. These verbal patterns aim to mentally penetrate the victim's mind in order to influence them, enabling the actor to obtain the desired information through implicit means. This paper presents an improved approach for detecting such implicit influential patterns. Furthermore, the proposed model is capable of identifying the specific locations of these influential elements within a conversation. To achieve this, the existing dataset was augmented using the reasoning capabilities of state-of-the-art language models. Our designed framework resulted in a 6% improvement in the detection of implicit influential patterns in conversations. Moreover, this approach improved the multi-label classification tasks related to both the techniques used for influence and the vulnerability of victims by 33% and 43%, respectively.

Md Kowsher, Nusrat Jahan Prottasha, Shiyun Xu et al.

Bidirectional language models have better context understanding and perform better than unidirectional models on natural language understanding tasks, yet the theoretical reasons behind this advantage remain unclear. In this work, we investigate this disparity through the lens of the Information Bottleneck (IB) principle, which formalizes a trade-off between compressing input information and preserving task-relevant content. We propose FlowNIB, a dynamic and scalable method for estimating mutual information during training that addresses key limitations of classical IB approaches, including computational intractability and fixed trade-off schedules. Theoretically, we show that bidirectional models retain more mutual information and exhibit higher effective dimensionality than unidirectional models. To support this, we present a generalized framework for measuring representational complexity and prove that bidirectional representations are strictly more informative under mild conditions. We further validate our findings through extensive experiments across multiple models and tasks using FlowNIB, revealing how information is encoded and compressed throughout training. Together, our work provides a principled explanation for the effectiveness of bidirectional architectures and introduces a practical tool for analyzing information flow in deep language models.

Mohammad Jahid Ibna Basher, Md Kowsher, Md Saiful Islam et al.

This paper introduces BnTTS (Bangla Text-To-Speech), the first framework for Bangla speaker adaptation-based TTS, designed to bridge the gap in Bangla speech synthesis using minimal training data. Building upon the XTTS architecture, our approach integrates Bangla into a multilingual TTS pipeline, with modifications to account for the phonetic and linguistic characteristics of the language. We pre-train BnTTS on 3.85k hours of Bangla speech dataset with corresponding text labels and evaluate performance in both zero-shot and few-shot settings on our proposed test dataset. Empirical evaluations in few-shot settings show that BnTTS significantly improves the naturalness, intelligibility, and speaker fidelity of synthesized Bangla speech. Compared to state-of-the-art Bangla TTS systems, BnTTS exhibits superior performance in Subjective Mean Opinion Score (SMOS), Naturalness, and Clarity metrics.

Md Kowsher, Abdul Rafae Khan, Jia Xu

In Chaos, a minor divergence between two initial conditions exhibits exponential amplification over time, leading to far-away outcomes, known as the butterfly effect. Thus, the distant future is full of uncertainty and hard to forecast. We introduce Group Reservoir Transformer to predict long-term events more accurately and robustly by overcoming two challenges in Chaos: (1) the extensive historical sequences and (2) the sensitivity to initial conditions. A reservoir is attached to a Transformer to efficiently handle arbitrarily long historical lengths, with an extension of a group of reservoirs to reduce the sensitivity to the initialization variations. Our architecture consistently outperforms state-of-the-art models in multivariate time series, including TimeLLM, GPT2TS, PatchTST, DLinear, TimeNet, and the baseline Transformer, with an error reduction of up to -59\% in various fields such as ETTh, ETTm, and air quality, demonstrating that an ensemble of butterfly learning can improve the adequacy and certainty of event prediction, despite of the traveling time to the unknown future.