Agha Ali Raza

Zohaib Khan, Muhammad Khaquan, Omer Tafveez et al.

The Transformer architecture has revolutionized deep learning through its Self-Attention mechanism, which effectively captures contextual information. However, the memory footprint of Self-Attention presents significant challenges for long-sequence tasks. Grouped Query Attention (GQA) addresses this issue by grouping queries and mean-pooling the corresponding key-value heads - reducing the number of overall parameters and memory requirements in a flexible manner without adversely compromising model accuracy. In this work, we introduce enhancements to GQA, focusing on two novel approaches that deviate from the static nature of grouping: Key-Distributed GQA (KDGQA) and Dynamic Key-Distributed GQA (DGQA), which leverage information from the norms of the key heads to inform query allocation. Specifically, KDGQA looks at the ratios of the norms of the key heads during each forward pass, while DGQA examines the ratios of the norms as they evolve through training. Additionally, we present Perturbed GQA (PGQA) as a case-study, which introduces variability in (static) group formation via subtracting noise from the attention maps. Our experiments with up-trained Vision Transformers, for Image Classification on datasets such as CIFAR-10, CIFAR-100, Food101, and Tiny ImageNet, demonstrate the promise of these variants in improving upon the original GQA through more informed and adaptive grouping mechanisms: specifically ViT-L experiences accuracy gains of up to 8% when utilizing DGQA in comparison to GQA and other variants. We further analyze the impact of the number of Key-Value Heads on performance, underscoring the importance of utilizing query-key affinities. Code is available on GitHub.

Samee Arif, Sualeha Farid, Abdul Hameed Azeemi et al.

This paper presents a novel methodology for generating synthetic Preference Optimization (PO) datasets using multi-model workflows. We evaluate the effectiveness and potential of these workflows in automating and enhancing the dataset generation process. PO dataset generation requires two modules: (1) $\textit{response evaluation}$, and (2) $\textit{response generation}$. In the $\textit{response evaluation}$ module, the responses from Large Language Models (LLMs) are evaluated and ranked - a task typically carried out by human annotators that we automate using LLMs. We assess the response evaluation module in a 2 step process. In step 1, we assess LLMs as evaluators using three distinct prompting strategies. In step 2, we apply the winning prompting strategy to compare the performance of LLM-as-a-Judge, LLMs-as-a-Jury, and LLM Debate. Our evaluation shows that GPT-4o-as-a-Judge is more consistent across all datasets. For the $\textit{response generation}$ module, we use the identified LLM evaluator configuration and compare different configurations of the LLM Feedback Loop. We use the win rate to determine the best multi-model configuration for generation. Experimenting with various configurations, we find that the LLM Feedback Loop, with Llama as the generator and Gemma as the reviewer, achieves a notable 71.8% and 73.8% win rate over single-model Llama and Gemma, respectively. After identifying the best configurations for both modules, we generate our PO datasets using the above pipeline.

Samee Arif, Abdul Hameed Azeemi, Agha Ali Raza et al.

In this paper, we compare general-purpose models, GPT-4-Turbo and Llama-3-8b, with special-purpose models--XLM-Roberta-large, mT5-large, and Llama-3-8b--that have been fine-tuned on specific tasks. We focus on seven classification and seven generation tasks to evaluate the performance of these models on Urdu language. Urdu has 70 million native speakers, yet it remains underrepresented in Natural Language Processing (NLP). Despite the frequent advancements in Large Language Models (LLMs), their performance in low-resource languages, including Urdu, still needs to be explored. We also conduct a human evaluation for the generation tasks and compare the results with the evaluations performed by GPT-4-Turbo, Llama-3-8b and Claude 3.5 Sonnet. We find that special-purpose models consistently outperform general-purpose models across various tasks. We also find that the evaluation done by GPT-4-Turbo for generation tasks aligns more closely with human evaluation compared to the evaluation the evaluation done by Llama-3-8b. This paper contributes to the NLP community by providing insights into the effectiveness of general and specific-purpose LLMs for low-resource languages.

Samee Arif, Sualeha Farid, Aamina Jamal Khan et al.

This paper presents a comprehensive evaluation of Urdu Automatic Speech Recognition (ASR) models. We analyze the performance of three ASR model families: Whisper, MMS, and Seamless-M4T using Word Error Rate (WER), along with a detailed examination of the most frequent wrong words and error types including insertions, deletions, and substitutions. Our analysis is conducted using two types of datasets, read speech and conversational speech. Notably, we present the first conversational speech dataset designed for benchmarking Urdu ASR models. We find that seamless-large outperforms other ASR models on the read speech dataset, while whisper-large performs best on the conversational speech dataset. Furthermore, this evaluation highlights the complexities of assessing ASR models for low-resource languages like Urdu using quantitative metrics alone and emphasizes the need for a robust Urdu text normalization system. Our findings contribute valuable insights for developing robust ASR systems for low-resource languages like Urdu.

Samee Arif, Muhammad Saad Haroon, Aamina Jamal Khan et al.

This paper introduces Kahaani, a multimodal, co-creative storytelling system that leverages Generative Artificial Intelligence, designed for children to address the challenge of sustaining engagement to foster educational narrative experiences. Here we define co-creative as a collaborative creative process in which both the child and Kahaani contribute to the generation of the story. The system combines Large Language Model (LLM), Text-to-Speech (TTS), Text-to-Music (TTM), and Text-to-Video (TTV) generation to produce a rich, immersive, and accessible storytelling experience. The system grounds the co-creation process in two classical storytelling framework, Freytag's Pyramid and Propp's Narrative Functions. The main goals of Kahaani are: (1) to help children improve their English skills, (2) to teach important life lessons through story morals, and (3) to help them understand how stories are structured, all in a fun and engaging way. We present evaluations for each AI component used, along with a user study involving three parent-child pairs to assess the overall experience and educational value of the system.

Abdul Hameed Azeemi, Ihsan Ayyub Qazi, Agha Ali Raza

Self-supervised speech recognition models require considerable labeled training data for learning high-fidelity representations for Automatic Speech Recognition (ASR) which is computationally demanding and time-consuming. We consider the task of identifying an optimal subset of data for efficient fine-tuning in self-supervised speech models for ASR. We discover that the dataset pruning strategies used in vision tasks for sampling the most informative examples do not perform better than random subset selection on fine-tuning self-supervised ASR. We then present the COWERAGE algorithm for representative subset selection in self-supervised ASR. COWERAGE is based on our finding that ensuring the coverage of examples based on training Word Error Rate (WER) in the early training epochs leads to better generalization performance. Extensive experiments with the wav2vec 2.0 and HuBERT model on TIMIT, Librispeech, and LJSpeech datasets show the effectiveness of COWERAGE and its transferability across models, with up to 17% relative WER improvement over existing dataset pruning methods and random sampling. We also demonstrate that the coverage of training instances in terms of WER values ensures the inclusion of phonemically diverse examples, leading to better test accuracy in self-supervised speech recognition models.

Shan Randhawa, Agha Ali Raza, Kentaro Toyama et al.

LLMs are increasingly being integrated into clinical workflows, yet they often lack clinical empathy, an essential aspect of effective doctor-patient communication. Existing NLP frameworks focus on reactively labeling empathy in doctors' responses but offer limited support for anticipatory modeling of empathy needs, especially in general health queries. We introduce the Empathy Applicability Framework (EAF), a theory-driven approach that classifies patient queries in terms of the applicability of emotional reactions and interpretations, based on clinical, contextual, and linguistic cues. We release a benchmark of real patient queries, dual-annotated by Humans and GPT-4o. In the subset with human consensus, we also observe substantial human-GPT alignment. To validate EAF, we train classifiers on human-labeled and GPT-only annotations to predict empathy applicability, achieving strong performance and outperforming the heuristic and zero-shot LLM baselines. Error analysis highlights persistent challenges: implicit distress, clinical-severity ambiguity, and contextual hardship, underscoring the need for multi-annotator modeling, clinician-in-the-loop calibration, and culturally diverse annotation. EAF provides a framework for identifying empathy needs before response generation, establishes a benchmark for anticipatory empathy modeling, and enables supporting empathetic communication in asynchronous healthcare.

Samee Arif, Sualeha Farid, Awais Athar et al.

This paper introduces UQA, a novel dataset for question answering and text comprehension in Urdu, a low-resource language with over 70 million native speakers. UQA is generated by translating the Stanford Question Answering Dataset (SQuAD2.0), a large-scale English QA dataset, using a technique called EATS (Enclose to Anchor, Translate, Seek), which preserves the answer spans in the translated context paragraphs. The paper describes the process of selecting and evaluating the best translation model among two candidates: Google Translator and Seamless M4T. The paper also benchmarks several state-of-the-art multilingual QA models on UQA, including mBERT, XLM-RoBERTa, and mT5, and reports promising results. For XLM-RoBERTa-XL, we have an F1 score of 85.99 and 74.56 EM. UQA is a valuable resource for developing and testing multilingual NLP systems for Urdu and for enhancing the cross-lingual transferability of existing models. Further, the paper demonstrates the effectiveness of EATS for creating high-quality datasets for other languages and domains. The UQA dataset and the code are publicly available at www.github.com/sameearif/UQA.

Samee Arif, Zohaib Khan, Maaidah Kaleem et al.

This paper presents a systematic analysis of biases in open-source Large Language Models (LLMs), across gender, religion, and race. Our study evaluates bias in smaller-scale Llama and Gemma models using the SALT ($\textbf{S}$ocial $\textbf{A}$ppropriateness in $\textbf{L}$LM-Generated $\textbf{T}$ext) dataset, which incorporates five distinct bias triggers: General Debate, Positioned Debate, Career Advice, Problem Solving, and CV Generation. To quantify bias, we measure win rates in General Debate and the assignment of negative roles in Positioned Debate. For real-world use cases, such as Career Advice, Problem Solving, and CV Generation, we anonymize the outputs to remove explicit demographic identifiers and use DeepSeek-R1 as an automated evaluator. We also address inherent biases in LLM-based evaluation, including evaluation bias, positional bias, and length bias, and validate our results through human evaluations. Our findings reveal consistent polarization across models, with certain demographic groups receiving systematically favorable or unfavorable treatment. By introducing SALT, we provide a comprehensive benchmark for bias analysis and underscore the need for robust bias mitigation strategies in the development of equitable AI systems.

Abdullah Hashmat, Muhammad Arham Mirza, Agha Ali Raza

With the widespread adoption of Large Language Models (LLMs) across various applications, it is empirical to ensure their fairness across all user communities. However, most LLMs are trained and evaluated on Western centric data, with little attention paid to low-resource languages and regional contexts. To address this gap, we introduce PakBBQ, a culturally and regionally adapted extension of the original Bias Benchmark for Question Answering (BBQ) dataset. PakBBQ comprises over 214 templates, 17180 QA pairs across 8 categories in both English and Urdu, covering eight bias dimensions including age, disability, appearance, gender, socio-economic status, religious, regional affiliation, and language formality that are relevant in Pakistan. We evaluate multiple multilingual LLMs under both ambiguous and explicitly disambiguated contexts, as well as negative versus non negative question framings. Our experiments reveal (i) an average accuracy gain of 12\% with disambiguation, (ii) consistently stronger counter bias behaviors in Urdu than in English, and (iii) marked framing effects that reduce stereotypical responses when questions are posed negatively. These findings highlight the importance of contextualized benchmarks and simple prompt engineering strategies for bias mitigation in low resource settings.

Abdul Hameed Azeemi, Ihsan Ayyub Qazi, Agha Ali Raza

Active learning (AL) techniques reduce labeling costs for training neural machine translation (NMT) models by selecting smaller representative subsets from unlabeled data for annotation. Diversity sampling techniques select heterogeneous instances, while uncertainty sampling methods select instances with the highest model uncertainty. Both approaches have limitations - diversity methods may extract varied but trivial examples, while uncertainty sampling can yield repetitive, uninformative instances. To bridge this gap, we propose Hybrid Uncertainty and Diversity Sampling (HUDS), an AL strategy for domain adaptation in NMT that combines uncertainty and diversity for sentence selection. HUDS computes uncertainty scores for unlabeled sentences and subsequently stratifies them. It then clusters sentence embeddings within each stratum and computes diversity scores by distance to the centroid. A weighted hybrid score that combines uncertainty and diversity is then used to select the top instances for annotation in each AL iteration. Experiments on multi-domain German-English and French-English datasets demonstrate the better performance of HUDS over other strong AL baselines. We analyze the sentence selection with HUDS and show that it prioritizes diverse instances having high model uncertainty for annotation in early AL iterations.

Haris Bin Zia, Agha Ali Raza, Awais Athar

State-of-the-art Natural Language Processing algorithms rely heavily on efficient word segmentation. Urdu is amongst languages for which word segmentation is a complex task as it exhibits space omission as well as space insertion issues. This is partly due to the Arabic script which although cursive in nature, consists of characters that have inherent joining and non-joining attributes regardless of word boundary. This paper presents a word segmentation system for Urdu which uses a Conditional Random Field sequence modeler with orthographic, linguistic and morphological features. Our proposed model automatically learns to predict white space as word boundary as well as Zero Width Non-Joiner (ZWNJ) as sub-word boundary. Using a manually annotated corpus, our model achieves F1 score of 0.97 for word boundary identification and 0.85 for sub-word boundary identification tasks. We have made our code and corpus publicly available to make our results reproducible.

Haris Bin Zia, Agha Ali Raza, Awais Athar

State-of-the-art speech recognition systems rely heavily on three basic components: an acoustic model, a pronunciation lexicon and a language model. To build these components, a researcher needs linguistic as well as technical expertise, which is a barrier in low-resource domains. Techniques to construct these three components without having expert domain knowledge are in great demand. Urdu, despite having millions of speakers all over the world, is a low-resource language in terms of standard publically available linguistic resources. In this paper, we present a grapheme-to-phoneme conversion tool for Urdu that generates a pronunciation lexicon in a form suitable for use with speech recognition systems from a list of Urdu words. The tool predicts the pronunciation of words using a LSTM-based model trained on a handcrafted expert lexicon of around 39,000 words and shows an accuracy of 64% upon internal evaluation. For external evaluation on a speech recognition task, we obtain a word error rate comparable to one achieved using a fully handcrafted expert lexicon.