United States

Generating contextually coherent multi-turn dialogue in Telugu requires resolving three deeply interacting constraints absent from generic LLM prompting: morphologically encoded social hierarchy (honorific verb conjugations), strict SOV agglutinative syntax, and culturally governed emotional logic formalised in Natyashastra rasa theory (Bharata Muni, 1951). We introduce LIMP (Linguistically-Informed Multi-Strategy Prompting), an inference-time, training-free framework that injects expert linguistic and cultural knowledge into prompt structure, requiring no fine-tuning or labelled data. We empirically evaluate two strategies on 10,000 stratified evaluation instances from the IndicDialogue Telugu corpus (Arnob et al., 2024): LIMP-RAW, a dense constraint prompt, and LIMP-COT, a six-stage analytical scaffold grounded in rasa theory and Telugu morphological grammar. Our primary finding is that LIMP-COT achieves approximately 2× higher morphosyntactic surface fidelity than LIMP-RAW on GEMMA-3-1B-IT (Gemma Team, Google DeepMind, 2025) (1B parameters): Jaccard = 0.0436 vs. 0.0211, Dice = 0.0792 vs. 0.0411 (p &lt; 0.001, Cohen’s d = 0.57), demonstrating that sequential analytical commitment to linguistic constraints produces more form-faithful Telugu than holistic constraint injection. Concurrently, LIMP-RAW achieves near-ceiling semantic fidelity (BERTSCORE F1 = 0.9709), exceeding both LIMP-COT (0.9637) and SARVAM-1 (Sarvam AI, 2024) (2B, Indic-pretrained; 0.9680) on this dimension. This semantic–lexical dissociation—no single configuration dominates across both metric classes—is itself a substantive finding: in agglutinative Telugu, semantic paraphrase fidelity and morphosyntactic surface fidelity are orthogonal evaluation dimensions. On lexical metrics specifically, LIMP-COT with a 1B general-purpose model surpasses SARVAM-1 under matched prompting (Jaccard = 0.0436 vs. 0.0052), suggesting that structured linguistic scaffolding is a stronger lever than parametric scale for form-faithful generation.

ACL 2026

LIMP: Linguistically-Informed Multi-Strategy Prompting for Telugu Multi-Turn Dialogue Generation

telugu dialogue generation

prompt engineering

low-resource languages

large language models

natural language processing

It is my great pleasure and honor to welcome you to the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), held in San Diego, California, USA, July 2–7, 2026. ACL 2026 continues our fields tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.

It is exciting to see the many developments in the field reflected at our conference, such as the advancement of large language models (born out of our community) into many application directions. This is not without challenges, partly due to unsolved problems, and partly due to misuse of the technology. All this is reflected well at ACL 2026. The conference also reflects that the field is not only about advancing technology, but also the science of fundamentally understanding language and how it is use by many different speaker communities. <br>
[Full Message from General Chair Philipp Koehn](https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing) <br><br><br>
<html><button style="display: inline-flex; align-items: center; justify-content: center; white-space: nowrap; border-radius: 9999px; font-weight: bold; background: #7c3aed; color: white; font-family: 'Space Grotesk', sans-serif; height: 40px; font-size: 16px; padding: 0 20px; border: none; cursor: pointer" onclick="window.open('https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing')">Conference Handbook</button></html>
<br><br><br><br>

To access this event page, you need to log in with the **email address you registered with**. <br>Access credentials will be sent to your email from Underline -  subject line "Welcome to ACL 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://2026.aclweb.org/registration/) first.

**Online Registration Form**: https://acl.swoogo.com/acl2026

Registration Required

The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026) will take place in San Diego, California, from July 2 - 7, 2026.

Generating contextually coherent multi-turn dialogue in Telugu requires resolving three deeply interacting constraints absent from generic LLM prompting: morphologically encoded social hierarchy (honorific verb conjugations), strict SOV agglutinative syntax, and culturally governed emotional logic formalised in Natyashastra rasa theory (Bharata Muni, 1951). We introduce LIMP (Linguistically-Informed Multi-Strategy Prompting), an inference-time, training-free framework that injects expert linguistic and cultural knowledge into prompt structure, requiring no fine-tuning or labelled data. We empirically evaluate two strategies on 10,000 stratified evaluation instances from the IndicDialogue Telugu corpus (Arnob et al., 2024): LIMP-RAW, a dense constraint prompt, and LIMP-COT, a six-stage analytical scaffold grounded in rasa theory and Telugu morphological grammar. Our primary finding is that LIMP-COT achieves approximately 2× higher morphosyntactic surface fidelity than LIMP-RAW on GEMMA-3-1B-IT (Gemma Team, Google DeepMind, 2025) (1B parameters): Jaccard = 0.0436 vs. 0.0211, Dice = 0.0792 vs. 0.0411 (p < 0.001, Cohen’s d = 0.57), demonstrating that sequential analytical commitment to linguistic constraints produces more form-faithful Telugu than holistic constraint injection. Concurrently, LIMP-RAW achieves near-ceiling semantic fidelity (BERTSCORE F1 = 0.9709), exceeding both LIMP-COT (0.9637) and SARVAM-1 (Sarvam AI, 2024) (2B, Indic-pretrained; 0.9680) on this dimension. This semantic–lexical dissociation—no single configuration dominates across both metric classes—is itself a substantive finding: in agglutinative Telugu, semantic paraphrase fidelity and morphosyntactic surface fidelity are orthogonal evaluation dimensions. On lexical metrics specifically, LIMP-COT with a 1B general-purpose model surpasses SARVAM-1 under matched prompting (Jaccard = 0.0436 vs. 0.0052), suggesting that structured linguistic scaffolding is a stronger lever than parametric scale for form-faithful generation.

workshop paper

In this digital era, the use of social media and online content in regional languages like Tamil is growing exponentially. Consequently, the volume of cyber abuse and gender based hate content is also on the rise. The presence of abusive language in regional languages such as Tamil poses a huge challenge. Tamil language has a highly inflected structure, and contains a large number of informal and colloquial words, dialects, and implicit and context dependent words. Therefore, there is a strong need for robust machine learning models to back up the efforts for online safety. We experimented various fine-tuned transformer-based pretrained language models which includes mBERT, MuRIL, XLM-RoBERTa, IndicBERT and L3Cube Tamil-BERT for the task of abusive content prediction. In this work, we investigate the effectiveness of transformer based models in abusive content detection in Tamil. Our comparative study shows that the transformer-based models are very effective in identifying the implicit, contextual and subtle language abusiveness present in Tamil. Experimental results demonstrate that Tamil-BERT (L3Cube) achieves a state-of-the-art accuracy of 80.72%. Its superior performance is attributed to exclusive training on Tamil corpora, providing the morphological precision necessary to navigate the language's structural complexities. The source code and trained models are available online.

DPR@DravidianLangTech 2026: Transformer-Based Abusive Content Detection for Tamil Text Targeting Women on Social Media

This paper presents a study of speech-to-speech translation for low-resource Dravidian languages, focusing on Tamil, Telugu, and Kannada. We investigate the efficacy of the Cascaded Modular system with the End-to-end system in both zero-shot and fine-tuned settings. The Cascaded Modular approach combines an ASR Module (Whisper-based ASR for English speech; IndicConformer for Dravidian speech), a Text-to-Text translation module (IndicTrans2), and a Speech synthesis module (Indic Parler-TTS), whereas SeamlessM4T is used as the End-to-end system. For parameter-efficient Low-Rank Adaptation (LoRA) fine-tuning to adapt the translation component to a domain-specific dataset, we use FLEURS and Mann-ki-Baat (a subset of BhasaAnuvaad dataset). Cascaded Modular systems achieve BLEU scores ranging from 3.17 to 19.18 in the zero-shot setting and 5.08 to 19.18 after fine-tuning, whereas the End-to-end model ranges from 3.02 to 15.72 in zero-shot settings across languages and 4.11 to 16.84 after fine-tuning. The results show that Cascaded Modular systems consistently outperform the End-to-end model across both setups. Note that parameter-efficient fine-tuning yields significant improvements in translation quality and speech generation performance for low-resource Dravidian speech translation.

Cascaded Modular or End-to-End? : An Investigation on Speech-to-Speech Translation Task for Dravidian Languages

Among various forms of online content, hope speech plays a crucial role in promoting positivity, encouragement, and social harmony, particularly in coe-mixed and culturally diverse communities. To address the challenges of building applications for Hope Speech Detection in Code- Mixed Tulu Language shared task organized by DravidianLangTech@ACL 2026, in this paper, we team - MUCS, describe the proposed methodology submitted for the shared task. As there are no pretrained models for Tulu, we explored multiple features - word n-grams (n = 1, 3), character n-grams (n = 1, 3), syllable n-grams (n = 1, 3) and sub-word features to train ensemble of classical Machine Learning (ML) models: i) Multinomial Naive Bayes and LR classifiers and ii) kNN and DT classifiers, both with soft-voting. Experimental results demonstrate that feature integration effectively captures lexical, sub-lexical, and phonological cues in noisy code-mixed text. The system achieves competitive performance on both development and test datasets, highlighting the effectiveness of feature-based approaches for low-resource hope speech detection in code-mixed Tulu.

MUCS@Dravidianlangtech@ACL2026: Hope Speech Detection in Code-Mixed Tulu Language Using Multiple Features

Dialectal variation poses a significant challenge to Automatic Speech Recognition (ASR), particularly for low resource morphologically rich languages such as Tamil. Although widely spoken in India, Sri Lanka, and the global diaspora, Tamil exhibits substantial phonetic, lexical, and prosodic variation across dialects, complicating both dialect classification and speech recognition. In this work, we address these tasks within a unified framework. We evaluate state-of-the-art models for dialect classification, including Whisper, CLDNN, wav2vec, and wavLM, and for ASR, Whisper and a zero-shot Conformer. Among them, Whisper achieves the best performance, obtaining a macro F1-score of 0.46 for dialect classification and a word error rate of 0.57 for ASR. These results highlight the strong generalization capability of transformer-based foundation models across dialects and languages. The code is publicly available in github for research purpose.

IIITK_SpeechScape@DravidianLangTech 2026: Dialect based speech recognition and classification using Speech Foundation Models and Deep Learning Techniques

Prompt recovery in large language models (LLMs) is the task of inferring the communicative intent and stylistic framing of the original instruction from model-generated output. This task is especially challenging for low-resource Dravidian languages such as Telugu, where agglutinative morphology, register variation, and scarce annotated data complicate stylistic modelling. In this paper, we present our system for the Shared Task on Prompt Recovery for LLM in Telugu at DravidianLangTech @ ACL 2026, which aims to classifiy Telugu transcript excerpts into nine communicative style categories: Formal, Informal, Optimistic, Pessimistic, Humorous, Serious, Inspiring, Authoritative, and Persuasive. We have implemented a transformer-based approach using ai4bharat/IndicBERTv2-MLM-only, MuRIL-base and Telugu-BERT for Telugu communicative style classification. Our system fine-tunes the pretrained Indic language training samples to capture stylistic patterns in Telugu transcripts. The training setup incorporates label smoothing, the AdamW optimizer with linear warmup and decay scheduling, and early stopping based on macro-averaged F1 to ensure stable convergence. Our approach achieved a macro F1 score of 0.2993 on the evaluation set, demonstrating the effectiveness of Indic-focused pretrained models for stylistic analysis in low-resource language settings.

Cuet Yet Another Baseline@DravidianLangTech 2026: Shared Task on Prompt Recovery for LLM in Telugu

Detecting abusive language in Tamil social media is a genuinely difficult problem. The language is morphologically rich, speakers routinely mix Tamil with English, and informal romanised Tamil is common enough to confuse models trained primarily on formal text. This work presents a system for binary classification of Tamil comments into Abusive and Non-Abusive categories, submitted to the DravidianLangTech@ACL 2026 shared task. MuRIL, a BERT-based encoder pre-trained on 17 Indian languages and their transliterated equivalents, is fine-tuned, and it is shown that this Indian-language-specific pre-training provides a meaningful advantage over generic multilingual baselines. The system achieves a macro-averaged F1 of 0.83 on the validation set, compared to 0.79 for XLM-RoBERTa and 0.77 for mBERT under identical training conditions, establishing a strong transformer-based baseline for abusive language detection in code-mixed Tamil.

PrimeLine@DravidianLangTech 2026: Abusive Tamil Comment Detection Using MuRIL

Hope speech detection in low-resource, code-mixed languages presents a genuine challenge for natural language processing. Tulu, a Dravidian language spoken along the coastal regions of Karnataka and Kerala, is one such language where social media content is deeply code-mixed, blending Tulu, Kannada script, and English within a single comment. Two classification tasks are addressed: a four-class coarse-grained setting (Track 1) and a five-class fine-grained setting (Track 2). XLM-RoBERTa, a cross-lingual transformer pre-trained on more than 100 languages, is fine-tuned on the task-provided datasets using Google Colab with an NVIDIA T4 GPU. The system achieves a Macro F1-score of 0.34 on Track 1 and 0.19 on Track 2 on the official Codabench evaluation, establishing the first transformer-based baseline for hope speech classification in Tulu.

PrimeLine@DravidianLangTech 2026: Hope Speech Detection in Tulu Using XLM-RoBERTa for Coarse and Fine-Grained Classification

An increased amount of violence against women has resulted from social media. Automatic detection of Tamil text that is abusive toward women is now an important area for research in natural language processing. This paper describes an experimental comparison between several classification algorithms, including support vector machine (SVM), decision tree, and random forest, to classify Tamil-language text as abusive or non-abusive. The data set used in evaluating the models contains Tamil-language social media text that is gender-targeted and abusive. The results show clearly that transformer models outperform traditional machine-learning models. In particular, RoBERTa had an accuracy rate of 82.35% for abusive language detection, suggesting that transformer models effectively capture contextual and semantic features of Tamil and can thus be applied to abuse detection across all low-resource Dravidian languages.

VISION@DravidianLangTech 2026: Abusive Tamil Text Detection Targeting Women on Social Media

Hope speech detection is an important task in understanding emotionally constructive communication in online platforms, especially in low-resource and code-mixed languages. This paper describes our system submitted to the first shared task on Hope Speech Detection in Code-Mixed Tulu, organized by DravidianLangTech@ACL 2026. The shared task consists of two tasks: Task 1 - Coarse-Grained Hope Tone Classification and Task 2 - Fine-Grained Hope Type Classification, with the objective of detecting and classifying the tone and type of hope expressed in code-mixed Tulu texts. We experimented with Logistic Regression (LR) and Linear Support Vector Classifier (LinearSVC) - classical Machine Learning (ML) approaches, trained with Term Frequency and Inverse Document Frequency (TF-IDF) of word ngrams (n = 1, 2). For Task 1, we employed both models, whereas for Task 2, we employed only the LR model. Linear SVC obtained a macro F1-score of 0.51 in Task 1 and secured 4th rank, while the LR model obtained a macro F1-score of 0.37 in Task 2 and secured 5th rank. The results demonstrate that traditional ML approaches remain effective for low-resource code-mixed language scenarios.

SJM_MINDS@DravidianLangTech@ACL2026: Machine Learning Approaches for Hope Speech Detection in Code-Mixed Tulu

This paper describes the system that our Still-Loading team designed to run the Telugu Prompt-Style Recovery shared task at DravidianLangTech@ACL 2026. The purpose of the given task is categorizing Telugu transcript passages as belonging to one of 9 communicative styles: Formal, Informal, Optimistic, Pessimistic, Humorous, Serious, Inspiring, Authoritative, and Persuasive. We compared several multilingual Transformer-based models, i.e. MuRIL, XLM-RoBERTa-Large, mBERT, and IndicBERTv2. We chose a "Turbo Sandwich" preprocessing strategy which helps to give more emphasis to lexical deltas, in addition to Focal Loss. Our system based on the MuRIL was rated at the 7th place in the official leaderboard with a Macro-F1 rating of 0.1703. The source code to reproduce our experiments is publicly available on Still-Loading-Prompt-Recovery-for-LLM-in-Telugu (https://github.com/Priyontee1713/Still-Loading-Prompt-Recovery-for-LLM-in-Telugu).

Downloads

Next from ACL 2026

DPR@DravidianLangTech 2026: Transformer-Based Abusive Content Detection for Tamil Text Targeting Women on Social Media

Similar lecture

Cuet Yet Another Baseline@DravidianLangTech 2026: Shared Task on Prompt Recovery for LLM in Telugu

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from ACL 2026

DPR@DravidianLangTech 2026: Transformer-Based Abusive Content Detection for Tamil Text Targeting Women on Social Media

Similar lecture

Cuet Yet Another Baseline@DravidianLangTech 2026: Shared Task on Prompt Recovery for LLM in Telugu

Downloads