Abstract

Current computational approaches for analysing or generating code-mixed sentences do not explicitly model ``naturalness'' or ``acceptability'' of code-mixed sentences, but rely on training corpora to reflect distribution of acceptable code-mixed sentences. Modelling human judgement for the acceptability of code-mixed text can help in distinguishing natural code-mixed text and enable quality-controlled generation of code-mixed text. To this end, we construct Cline - a dataset containing human acceptability judgements for English-Hindi~(en-hi) code-mixed text. Cline is the largest of its kind with 16,642 sentences, consisting of samples sourced from two sources: synthetically generated code-mixed text and samples collected from online social media. Our analysis establishes that popular code-mixing metrics such as CMI, Number of Switch Points, Burstines, which are used to filter/curate/compare code-mixed corpora have low correlation with human acceptability judgements, underlining the necessity of our dataset. Experiments using Cline demonstrate that simple Multilayer Perceptron (MLP) models when trained solely using code-mixing metrics as features are outperformed by fine-tuned pre-trained Multilingual Large Language Models (MLLMs). Specifically, among Encoder models XLM-Roberta and Bernice outperform IndicBERT across different configurations. Among Encoder-Decoder models, mBART performs better than mT5, however Encoder-Decoder models are not able to outperform Encoder-only models. Decoder-only models perform the best when compared to all other MLLMS, with Llama 3.2 - 3B models outperforming similarly sized Qwen, Phi models. Comparison with zero and fewshot capabilitites of ChatGPT show that MLLMs fine-tuned on larger data outperform ChatGPT, providing scope for improvement in code-mixed tasks. Zero-shot transfer from En-Hi to En-Te acceptability judgments are better than random baselines.

Abstract

Graph Neural Networks (GNNs) are increasingly being used for a variety of ML applications on graph data. Because graph data does not follow the independently and identically distributed i.i.d. assumption, adversarial manipulations or incorrect data can propagate to other data points through message passing, which deteriorates the model's performance. To allow model developers to remove the adverse effects of manipulated entities from a trained GNN, we study the recently formulated problem of Corrective Unlearning. We find that current graph unlearning methods fail to unlearn the effect of manipulations even when the whole manipulated set is known. We introduce a new graph unlearning method,Cognac, which can unlearn the effect of the manipulation set even when only % of it is identified. It recovers most of the performance of a strong oracle with fully corrected training data, even beating retraining from scratch without the deletion set, and is 8x more efficient while also scaling to large datasets. We hope our work assists GNN developers in mitigating harmful effects caused by issues in real-world data, post-training.

Abstract

There is growing excitement about the potential of Language Models (LMs) to accelerate scientific discovery. Falsifying hypotheses is key to scientific progress, as it allows claims to be iteratively refined over time. This process requires significant researcher effort, reasoning, and ingenuity. Yet current benchmarks for LMs predominantly assess their ability to generate solutions rather than challenge them. We advocate for developing benchmarks that evaluate this inverse capability — creating counterexamples for subtly incorrect solutions. To demonstrate this approach, we start with the domain of algorithmic problem solving, where counterexamples can be evaluated automatically using code execution. Specifically, we introduce REFUTE, a dynamically updating benchmark that includes recent problems and incorrect submissions from programming competitions, where human experts successfully identified counterexamples. Our analysis finds that the best reasoning agents, even OpenAI o3-mini (high) with code execution feedback, can create counterexamples for only < 9% of incorrect solutions in REFUTE, even though ratings indicate its ability to solve up to 48% of these problems from scratch. We hope our work spurs progress in evaluating and enhancing LMs’ ability to falsify incorrect solutions — a capability that is crucial for both accelerating research and making models self-improve through reliable reflective reasoning.

Abstract

Understanding user emotions to identify user opinion, sentiment, stance, and preferences has become a hot topic of research in the last few years. Many studies and datasets are designed for user emotion analysis including news websites, blogs, and user tweets. However, there is little exploration of political emotions in the Indian context for multi-label emotion detection. This paper presents a PoliEMO dataset—a novel benchmark corpus of political tweets in a multi-label setup for Indian elections, consisting of over 3512 tweets manually annotated. In this work, 6792 labels were generated for six emotion categories: anger, insult, joy, neutral, sadness, and shameful. Next, PoliEMO dataset is used to understand emotions in a multi-label context using state-of-the-art machine learning algorithms with multi-label classifier (binary relevance (BR), label powerset (LP), classifier chain (CC), and multi-label k-nearest neighbors (MkNN)) and deep learning models like convolutional neural network (CNN), long short-term memory (LSTM), bidirectional long short-term memory (Bi-LSTM), and transfer learning model, i.e., bidirectional encoder representations from transformers (BERT). Experiments and results show Bi-LSTM performs better with micro-averaged F1 score of 0.81, macro-averaged F1 score of 0.78, and accuracy 0.68 as compared to state-of-the-art approaches.

Abstract

In covering elections, journalists often use conflict frames which depict events and issues as adversarial, often highlighting confronta- tions between opposing parties. Although conflict frames result in more citizen engagement, they may distract from substantive policy discussion. In this work, we analyze the use of conflict frames in online English-language news articles by seven major news outlets in the 2014 and 2019 Indian general elections. We find that the use of conflict frames is not linked to the news outlets’ ideological biases but is associated with TV-based (rather than print-based) media. Further, the majority of news outlets do not exhibit ideolog- ical biases in portraying parties as aggressors or targets in articles with conflict frames. Finally, comparing news articles reporting on political speeches to their original speech transcripts, we find that, on average, news outlets tend to consistently report on attacks on the opposition party in the speeches but under-report on more substantive electoral issues covered in the speeches such as farmers’ issues and infrastructure.

Abstract

Engaging in political discussions is crucial in democratic societies, yet many individuals remain politically disinclined due to various factors such as perceived knowledge gaps, conflict avoidance, or a sense of disconnection from the political system. In this paper, we explore the potential of personal narratives—short, first-person ac- counts emphasizing personal experiences—as a means to empower these individuals to participate in online political discussions. Using a text classifier that identifies personal narratives, we conducted a large-scale computational analysis to evaluate the relationship between the use of personal narratives and participation in po- litical discussions on Reddit. We find that politically disinclined individuals (PDIs) are more likely to use personal narratives than more politically active users. Personal narratives are more likely to attract and retain politically disinclined individuals in political discussions than other comments. Importantly, personal narratives posted by politically disinclined individuals are received more posi- tively than their other comments in political communities. These results emphasize the value of personal narratives in promoting inclusive political discourse.

Abstract

Large Language Models (LLMs) often inherit biases from the web data they are trained on, which contains stereotypes and prejudices. Current methods for evaluating and mitigating these biases rely on bias-benchmark datasets. These benchmarks measure bias by observing an LLM’s behavior on biased statements. However, these statements lack contextual considerations of the situations they try to present. To address this, we introduce a contextual reliability framework, which evaluates model robustness to biased statements by considering the various contexts in which they may appear. We develop the Context-Oriented Bias Indicator and Assessment Score (COBIAS) to measure a biased statement’s reliability in detecting bias, based on the variance in model behavior across different contexts. To evaluate the metric, we augmented 2,291 stereotyped statements from two existing benchmark datasets by adding contextual information. We show that COBIAS aligns with human judgment on the contextual reliability of biased statements (Spearman’s 𝜌 = 0.65, 𝑝 = 3.4 ∗ 10−60) and can be used to create reliable benchmarks, which would assist bias mitigation works.

Abstract

Graph Neural Networks (GNNs) have emerged as powerful tools for learning representations of graph-structured data, demonstrating remarkable performance across various tasks. Recognising their importance, there has been extensive research focused on explaining GNN predictions, aiming to enhance their interpretability and trustworthiness. However, GNNs and their explainers face a notable challenge: graphs are primarily designed to model pair-wise relationships between nodes, which can make it tough to capture higher-order, multi-node interactions. This characteristic can pose difficulties for existing explainers in fully representing multi-node relationships. To address this gap, we present Framework For Higher-Order Representations In Graph Explanations (FORGE), a framework that enables graph explainers to capture such interactions by incorporating higher-order structures, resulting in more accurate and faithful explanations. Extensive evaluation shows that on average real-world datasets from the GraphXAI benchmark and synthetic datasets across various graph explainers, FORGE improves average explanation accuracy by 1.9x and 2.25x, respectively. We perform ablation studies to confirm the importance of higher-order relations in improving explanations, while our scalability analysis demonstrates FORGE's efficacy on large graphs.

Abstract

Knowledge Graph Completion (KGC) aims to predict missing entities or relations in knowledge graph but it becomes computationally expensive as KG scales. Existing research focuses on bilinear pooling-based factorization methods (LowFER, TuckER) to solve this problem. These approaches introduce too many trainable parameters which obstruct the deployment of these techniques in many real-world scenarios. In this paper, we introduce a novel parameter-efficient framework, KGRP which a) approximates bilinear pooling using Kernelized Random Projection matrix b) employs CNN for the better fusion of entities and relations to infer missing links. Our experimental results show that KGRP has 73% fewer parameters as compared to the state-of-theart approaches (LowFER, TuckER) for the knowledge graph completion task while retaining 88% performance for the best baseline. Furthermore, we also provide novel insights on the interpretability of relation embeddings. We also test the effectiveness of KGRP on a large-scale recruitment knowledge graph of 0.25 M entities.

Abstract

Graph neural networks (GNNs) are increasingly being used on sensitive graph-structured data, necessitating techniques for handling unlearning requests on the trained models, particularly node unlearning. However, unlearning nodes on GNNs is challenging due to the interdependence between the nodes in a graph. We compare MEGU, a state-of-the-art graph unlearning method, and SCRUB, a general unlearning method for classification, to investigate the efficacy of graph unlearning methods over traditional unlearning methods. Surprisingly, we find that SCRUB performs comparably or better than MEGU on random node removal and on removing an adversarial node injection attack. Our results suggest that 1) graph unlearning studies should incorporate general unlearning methods like SCRUB as baselines, and 2) there is a need for more rigorous behavioral evaluations that reveal the differential advantages of proposed graph unlearning methods. Our work, therefore, motivates future research into more comprehensive evaluations for assessing the true utility of graph unlearning algorithms.