Sarana Nutanong — accepted AI-conference papers

2025

NitiBench: Benchmarking LLM Frameworks on Thai Legal Question Answering Capabilities

Pawitsapak Akarajaradwong, Pirat Pothavorn, Chompakorn Chaksangchaichot, Panuthep Tasawong, Thitiwat Nopparatbundit, Keerakiat Pratai +1

EMNLP 2025

Large language models (LLMs) show promise in legal question answering (QA), yet Thai legal QA systems face challenges due to limited data and complex legal structures. We introduce NitiBench, a novel benchmark featuring two datasets: (1) NitiBench-CCL, covering Thai financial laws, and (2) NitiBench

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

Prior Prompt Engineering for Reinforcement Fine-Tuning

Pittawat Taveekitworachai, Potsawee Manakul, Sarana Nutanong, Kunat Pipatanakul

EMNLP 2025

This paper investigates prior prompt engineering (pPE) in the context of reinforcement fine-tuning (RFT), where language models (LMs) are incentivized to exhibit behaviors that maximize performance through reward signals. While existing RFT research has primarily focused on algorithms, reward shapin

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2025

Towards Better Understanding of Program-of-Thought Reasoning in Cross-Lingual and Multilingual Environments

Patomporn Payoungkhamdee, Pume Tuchinda, Jinheon Baek, Samuel Cahyawijaya, Can Udomcharoenchaikit, Potsawee Manakul +3

ACL 2025finding

Multi-step reasoning is essential for large language models (LLMs), yet multilingual performance remains challenging. While Chain-of-Thought (CoT) prompting improves reasoning, it struggles with non-English languages due to the entanglement of reasoning and execution. Program-of-Thought (PoT) prompt…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2025

WangchanThaiInstruct: An instruction-following Dataset for Culture-Aware, Multitask, and Multi-domain Evaluation in Thai

Peerat Limkonchotiwat, Pume Tuchinda, Lalita Lowphansirikul, Surapon Nonesung, Panuthep Tasawong, Alham Fikri Aji +2

EMNLP 2025

Large language models excel at instruction-following in English, but their performance in low-resource languages like Thai remains underexplored. Existing benchmarks often rely on translations, missing cultural and domain-specific nuances needed for real-world use. We present WangchanThaiInstruct, a

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2024

An Empirical Study of Multilingual Reasoning Distillation for Question Answering

Patomporn Payoungkhamdee, Peerat Limkonchotiwat, Jinheon Baek, Potsawee Manakul, Can Udomcharoenchaikit, Ekapol Chuangsuwanich +1

EMNLP 2024main

Reasoning is one crucial capability in Large Language Models (LLMs), allowing them to perform complex tasks such as solving math problems and multi-step planning. While reasoning capability can emerge in larger models, smaller ones usually have to rely on distillation to transfer this capability fro…

Cited by 2 Source ↗PDF ↗code ↗Scholar ↗

2024

Efficient Overshadowed Entity Disambiguation by Mitigating Shortcut Learning

Panuthep Tasawong, Peerat Limkonchotiwat, Potsawee Manakul, Can Udomcharoenchaikit, Ekapol Chuangsuwanich, Sarana Nutanong

EMNLP 2024main

Entity disambiguation (ED) is crucial in natural language processing (NLP) for tasks such as question-answering and information extraction. A major challenge in ED is handling overshadowed entities—uncommon entities sharing mention surfaces with common entities. The current approach to enhance perfo…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2024

Identifying and Mitigating Annotation Bias in Natural Language Understanding using Causal Mediation Analysis

Sitiporn Sae Lim, Can Udomcharoenchaikit, Peerat Limkonchotiwat, Ekapol Chuangsuwanich, Sarana Nutanong

ACL 2024findings

NLU models have achieved promising results on standard benchmarks. Despite state-of-the-art accuracy, analysis reveals that many models make predictions using annotation bias rather than the properties we intend the model to learn. Consequently, these models perform poorly on out-of-distribution dat…

Cited by 0 Source ↗PDF ↗Scholar ↗

2024

MIST: Mutual Information Maximization for Short Text Clustering

Krissanee Kamthawee, Can Udomcharoenchaikit, Sarana Nutanong

ACL 2024long

Short text clustering poses substantial challenges due to the limited amount of information provided by each text sample. Previous efforts based on dense representations are still inadequate as texts are not sufficiently segregated in the embedding space before clustering. Even though the state-of-t…

Cited by 2 Source ↗PDF ↗code ↗Scholar ↗

2024

McCrolin: Multi-consistency Cross-lingual Training for Retrieval Question Answering

Peerat Limkonchotiwat, Wuttikorn Ponwitayarat, Lalita Lowphansirikul, Potsawee Manakul, Can Udomcharoenchaikit, Ekapol Chuangsuwanich +1

EMNLP 2024finding

Automated question answering (QA) systems are increasingly relying on robust cross-lingual retrieval to identify and utilize information from multilingual sources, ensuring comprehensive and contextually accurate responses. Existing approaches often struggle with consistency across multiple language…

Cited by 5 Source ↗PDF ↗Scholar ↗

2024

Space Decomposition for Sentence Embedding

Wuttikorn Ponwitayarat, Peerat Limkonchotiwat, Ekapol Chuangsuwanich, Sarana Nutanong

ACL 2024findings

Determining sentence pair similarity is crucial for various NLP tasks. A common technique to address this is typically evaluated on a continuous semantic textual similarity scale from 0 to 5. However, based on a linguistic observation in STS annotation guidelines, we found that the score in the rang…

Cited by 1 Source ↗PDF ↗code ↗Scholar ↗

2023

Learning Geometric-Aware Properties in 2D Representation Using Lightweight CAD Models, or Zero Real 3D Pairs

Pattaramanee Arsomngern, Sarana Nutanong, Supasorn Suwajanakorn

CVPR 2023poster

Cross-modal training using 2D-3D paired datasets, such as those containing multi-view images and 3D scene scans, presents an effective way to enhance 2D scene understanding by introducing geometric and view-invariance priors into 2D features. However, the need for large-scale scene datasets can impe…

Cited by 3 Source ↗PDF ↗Scholar ↗

2023

Typo-Robust Representation Learning for Dense Retrieval

Panuthep Tasawong, Wuttikorn Ponwitayarat, Peerat Limkonchotiwat, Can Udomcharoenchaikit, Ekapol Chuangsuwanich, Sarana Nutanong

ACL 2023short

Dense retrieval is a basic building block of information retrieval applications. One of the main challenges of dense retrieval in real-world settings is the handling of queries containing misspelled words. A popular approach for handling misspelled queries is minimizing the representations discrepan…

Cited by 8 Source ↗PDF ↗code ↗Scholar ↗

2022

CL-ReLKT: Cross-lingual Language Knowledge Transfer for Multilingual Retrieval Question Answering

Peerat Limkonchotiwat, Wuttikorn Ponwitayarat, Can Udomcharoenchaikit, Ekapol Chuangsuwanich, Sarana Nutanong

NAACL 2022findings

Cross-Lingual Retrieval Question Answering (CL-ReQA) is concerned with retrieving answer documents or passages to a question written in a different language. A common approach to CL-ReQA is to create a multilingual sentence embedding space such that question-answer pairs across different languages a…

Cited by 9 Source ↗PDF ↗code ↗Scholar ↗

2022

ConGen: Unsupervised Control and Generalization Distillation For Sentence Representation

Peerat Limkonchotiwat, Wuttikorn Ponwitayarat, Lalita Lowphansirikul, Can Udomcharoenchaikit, Ekapol Chuangsuwanich, Sarana Nutanong

EMNLP 2022finding

Sentence representations are essential in many NLP tasks operating at the sentence level.Recently, research attention has shifted towards learning how to represent sentences without any annotations, i.e., unsupervised representation learning. Despite the benefit of training without supervised data,…

Cited by 8 Source ↗PDF ↗code ↗Scholar ↗

2022

Mitigating Spurious Correlation in Natural Language Understanding with Counterfactual Inference

Can Udomcharoenchaikit, Wuttikorn Ponwitayarat, Patomporn Payoungkhamdee, Kanruethai Masuk, Weerayut Buaphet, Ekapol Chuangsuwanich +1

EMNLP 2022main

Despite their promising results on standard benchmarks, NLU models are still prone to make predictions based on shortcuts caused by unintended bias in the dataset. For example, an NLI model may use lexical overlap as a shortcut to make entailment predictions due to repetitive data generation pattern…

Cited by 13 Source ↗PDF ↗Scholar ↗

2022

Thai Nested Named Entity Recognition Corpus

Weerayut Buaphet, Can Udomcharoenchaikit, Peerat Limkonchotiwat, Attapol Rutherford, Sarana Nutanong

ACL 2022findings

This paper presents the first Thai Nested Named Entity Recognition (N-NER) dataset. Thai N-NER consists of 264,798 mentions, 104 classes, and a maximum depth of 8 layers obtained from 4,894 documents in the domains of news articles and restaurant reviews. Our work, to the best of our knowledge, pres…

Cited by 6 Source ↗PDF ↗code ↗Scholar ↗

2022

Topic-Regularized Authorship Representation Learning

Jitkapat Sawatphol, Nonthakit Chaiwong, Can Udomcharoenchaikit, Sarana Nutanong

EMNLP 2022main

Authorship attribution is a task that aims to identify the author of a given piece of writing. We aim to develop a generalized solution that can handle a large number of texts from authors and topics unavailable in training data. Previous studies have proposed strategies to address only either unsee…

Cited by 5 Source ↗PDF ↗code ↗Scholar ↗

2021

Robust Fragment-Based Framework for Cross-lingual Sentence Retrieval

Nattapol Trijakwanich, Peerat Limkonchotiwat, Raheem Sarwar, Wannaphong Phatthiyaphaibun, Ekapol Chuangsuwanich, Sarana Nutanong

EMNLP 2021finding

Cross-lingual Sentence Retrieval (CLSR) aims at retrieving parallel sentence pairs that are translations of each other from a multilingual set of comparable documents. The retrieved parallel sentence pairs can be used in other downstream NLP tasks such as machine translation and cross-lingual word s…

Cited by 8 Source ↗PDF ↗Scholar ↗