Research

Papers are roughly grouped by topics and sorted based on date. The full list of articles is available on my Google Scholar.

Agentic Systems

MiniWeb overview: 65 connected websites across 15 domains for practicing and evaluating web-agent skills
Back to Basics: Drills Before Games for Web Agents
Under review, 2026
Minh Pham-Dinh, Md Farhan Ishmam, Hernan Yupanqui, Reaz Hassan Joarder, Kenneth Marino
Project Page Code Dataset
WebKD method overview: teacher feedback on student rollouts for web-agent distillation
Teaching Agents in their Own Style: Blackbox Teacher Distillation for Web Agents
ArXiv'26
Md Farhan Ishmam, Kenneth Marino
Project Page Paper Code
WebMix method overview: Scout, mixture hypernetwork, and adapted web agent
Don't Retrain, Remix: Adaptation of Web Agents using Mixture Hypernetworks
ArXiv'26
Md Farhan Ishmam, Minh Pham Dinh, Kenneth Marino
Project Page Paper Code
TimeWarp paper figure
TimeWarp: Evaluating Web Agents by Revisiting the Past
NeurIPS'26 Evaluations and Datasets
Md Farhan Ishmam, Kenneth Marino
Project Page Paper Code Dataset Models
DORA Explorer method overview: ranking and sampling candidate actions improves exploration and avoids repeated action loops
DORA Explorer: Improving the Exploration Ability of LLMs Without Training
ArXiv'26
Priya Gurjar, Md Farhan Ishmam, Kenneth Marino
Project Page Paper Code
KNOWS benchmark overview: a web agent creates a slide deck, evaluated with detailed content and visual checkpoints
The Hard Part Comes After Search: Benchmarking Web Agents on Synthesizing, Organizing, and Displaying Knowledge
EMNLP'26 Findings
Alexander Gill, Md Farhan Ishmam, Xuyen Nguyen, Neha Bhat, Parker Henry DeYoung, Fateme Hashemi Chaleshtori, Nathan Stringham, Kenneth Marino, Ana Marasović
Project Page Paper Code Dataset
Computer Use Survey paper figure
Computer Use Survey - A Visual Survey of Computer Use Agents
ICLR'26 Blogposts
Kenneth Marino, Md Farhan Ishmam, and Ana Marasovic
Blog

LLM Creativity

Nonlinear narrative example comparing chronological and telling order, showing how a later reveal explains an earlier event
Can LLMs Tell Stories Out of Order? A Corpus and Benchmark for Nonlinear Narrative Generation
ArXiv'26
Syed Rifat Raiyan, Md. Tariquzzaman, Md Farhan Ishmam, Tenghao Huang
Project Page Paper Code

Vision-Language Adapters

R-MMA paper figure
R-MMA: Enhancing Vision-Language Models with Recurrent Adapters for Few-Shot and Cross-Domain Generalization
WACV'26
Md Fahim*, Md Farhan Ishmam*, Mir Sazzat Hossain, M Ashraful Amin, Amin Ahsan Ali, AKM Mahbubur Rahman
Paper Supp Code Poster

Prompt Compression

FrugalPrompt prompt compression paper figure
FrugalPrompt: Reducing Contextual Overhead in Large Language Models via Token Attribution
ArXiv'26
Syed Rifat Raiyan, Md Farhan Ishmam, Abdullah Al Imran, Mohammad Ali Moni
Project Page Paper Code

Visual Question Answering (VQA)

Vision-language corruption robustness paper figure
Enhancing Vision Language Corruption Robustness using Cross Distribution & Prompted Denoisers
WACV'26
Sameer Shafayet Latif*, Sadab Shiper*, K. M. Rahiduzzaman Kiran*, Md Farhan Ishmam*, Md Azam Hossain, Abu Raihan Mostofa Kamal, Md Hamjajul Ashmafee
Paper Supp Code Poster Video
BanglaProtha dataset paper figure
BanglaProtha: Evaluating Vision Language Models in Underrepresented Long-tail Cultural Contexts
WACV'26
Md Fahim*, Md Sakib Ul Rahman*, Akm Moshiur Rahman*, Md Farhan Ishmam*, Md Tasmim Rahman, Fariha Tanjim Shifat, Fabiha Haider, Md Farhad Alam Bhuiyan
Paper Supp Code Dataset Poster
Visual Robustness Benchmark for VQA paper figure
Visual Robustness Benchmark for Visual Question Answering (VQA)
WACV'25
Md Farhan Ishmam*, Ishmam Tashdeed*, Talukder Asir Saadat*, Md Hamjajul Ashmafee, Abu Raihan Mostofa Kamal, Md Azam Hossain
Paper Supp Code Poster Slides Video
ChitroJera VQA dataset paper figure
ChitroJera: A Regionally Relevant Visual Question Answering Dataset for Bangla
ECML-PKDD'25
Deeparghya Dutta Barua*, Md Sakib Ul Rahman Sourove*, Md Farhan Ishmam*, Fabiha Haider, Fariha Tanjim Shifat, Md Fahim, Farhad Alam Bhuiyan
Paper Code Dataset Poster Slides
From Image to Language VQA survey paper figure
From Image to Language: A Critical Analysis of Visual Question Answering (VQA) Approaches, Challenges, and Opportunities
Information Fusion Journal | 2024
Md Farhan Ishmam, Md Sakib Hossain Shovon, Muhammad Firoz Mridha, Nilanjan Dey
Paper ArXiv

Transliteration and Code-Mixing

Robustness of LLMs to Transliteration Perturbations in Bangla paper figure
Robustness of LLMs to Transliteration Perturbations in Bangla
🏆 Best Paper BLP@IJCNLP-AACL'25
Fabiha Haider, Md Farhan Ishmam, Fariha Tanjim Shifat, Md Tasmim Rahman, Md Fahim, Md Farhad Alam Bhuiyan
Paper Code Poster Slides Video
BanglaTLit back-transliteration dataset paper figure
BanglaTLit: A Benchmark Dataset for Back-Transliteration of Romanized Bangla
EMNLP'24 Findings
Md Fahim*, Fariha Tanjim Shifat*, Fabiha Haider*, Deeparghya Dutta Barua, Md Sakib Ul Rahman Sourove, Md Farhan Ishmam, Farhad Alam Bhuiyan
Paper Code Dataset Poster Slides Video
BanTH hate speech dataset paper figure
BanTH: A Multi-label Hate Speech Detection Dataset for Transliterated Bangla
NAACL'25 Findings
Fabiha Haider*, Fariha Tanjim Shifat*, Md Farhan Ishmam*, Deeparghya Dutta Barua, Md Sakib Ul Rahman Sourove, Md Fahim, Farhad Alam Bhuiyan
ArXiv Code
BnSentMix sentiment analysis dataset paper figure
BnSentMix: A Diverse Bengali-English Code-Mixed Dataset for Sentiment Analysis
LoResLM@COLING'25
Sadia Alam, Md Farhan Ishmam, Navid Hasin Alvee, Md Shahnewaz Siddique, Md Azam Hossain, Abu Raihan Mostofa Kamal
Paper Code Dataset Poster Slides

Accessibility Technology

Prompting with Sign Language paper figure
Prompting with Sign Parameters for Low-resource Sign Language Instruction Generation
CV4A11y@ICCV'25
Md. Tariquzzaman, Md Farhan Ishmam, Saiyma Sittul Muna, Md Kamrul Hasan, Hasan Mahmud
Paper Code Dataset Video

General Text Classification

FourierKAN text classification paper figure
FourierKAN outperforms MLP on Text Classification Head Fine-tuning
FITML@NeurIPS'24
Abdullah Al Imran*, Md Farhan Ishmam*
Paper Code

Harmful Content

Penta NLP at EXIST 2024 paper figure
Penta NLP at EXIST 2024 Task 1–3: Sexism Identification, Source Intention, Sexism Categorization In Tweets
EXIST@CLEF'24
Fariha Tanjim Shifat, Fabiha Haider, Md Sakib Ul Rahman Sourove, Deeparghya Dutta Barua, Md Farhan Ishmam, Md Fahim, Farhad Alam Bhuiyan
Paper Code
Penta ML at EXIST 2024 paper figure
Penta ML at EXIST 2024: Tagging Sexism in Online Multimodal Content With Attention-enhanced Modal Context
EXIST@CLEF'24
Deeparghya Dutta Barua, Md Sakib Ul Rah Rahman Sourove, Fabiha Haider, Fariha Tanjim Shifat, Md Farhan Ishmam, Md Fahim, Farhad Alam Bhuiyan
Paper Code