Language models are increasingly deployed in interactive settings where users reason about facts over time rather than in isolation.In such scenarios, correct behavior requires models to maintain and update implicit temporal assumptions established earlier in a conversation.We study this challenge through the lens of temporal scope stability: the ability to preserve, override, or transfer time-scoped factual context across dialogue turns.We introduce ChronoScope, a large-scale diagnostic benchmark designed to isolate temporal scope behavior in controlled multi-turn interactions, comprising over one million deterministically generated question chains grounded in Wikidata.ChronoScope evaluates whether models can correctly retain inferred temporal scope when follow-up questions omit explicit time references, spanning implicit carryover, explicit scope switching, cross-entity transfer, and longer temporal trajectories.Through extensive evaluation of state-of-the-art language models, we find that temporal scope stability is frequently violated in controlled multiturn settings, with models often drifting toward present-day assumptions despite correct underlying knowledge.These failures intensify with interaction length and persist even under oracle context conditions, revealing a gap between single-turn factual accuracy and coherent temporal reasoning under sequential interaction.We make our dataset and evaluation suite publicly available at https://github. com/yashkumaratri/ChronoScope.

Multi-document Summarization (MDS) characterizes compressing information from multiple source documents to its succinct summary.An ideal summary should encompass all topics and accurately model cross-document relations expounded upon in the source documents.However, existing systems either impose constraints on the length of tokens during the encoding or falter in capturing the intricate cross-document relationships.These limitations impel the systems to produce summaries that are non-factual and unfaithful, thereby imparting an unfair comprehension of the topic to the readers.To counter these limitations and promote the information equivalence between the source document and generated summary, we propose FABRIC, a novel encoder-decoder model that uses pre-trained BART to comprehensively analyze linguistic nuances, simplicial complex layer to apprehend inherent properties that transcend pairwise associations and sheaf graph attention to effectively capture the heterophilic properties.We benchmark FABRIC with eleven baselines over four widely-used MDS datasets -Multinews, CQASumm, DUC and Opinosis, and show that FABRIC achieves consistent performance improvement across all the evaluation metrics (syntactical, semantical and faithfulness).We corroborate these improvements further through qualitative human evaluation.

The realm of scientific text summarization has experienced remarkable progress due to the availability of annotated brief summaries and ample data. However, the utilization of multiple input modalities, such as videos and audio, has yet to be thoroughly explored. At present, scientific multimodal-input-based text summarization systems tend to employ longer target summaries like abstracts, leading to an underwhelming performance in the task of text summarization.

Abstractive text summarization systems using recently improved RNN-based sequence-to-sequence architecture have shown great promise for single-document summarization. However, such neural models fail to perpetuate the performance in the multi-document summarization setting owing to the long-range dependencies within the documents, overlapping/contradicting facts and extrinsic model hallucinations. These shortcomings augment the model to generate inconsistent, repetitive and non-factual summaries. In this work, we introduce REISA, a sequence-to-sequence model with a novelreinforced selective attention spanthat attends over the input and recalibrates the local attention weights to focus on important segments while generating output at each time step. REISA utilizes a reinforcement learning-based policy gradient algorithm to reward the model and formulate attention distributions over the encoder input. We further benchmark REISA on two widely-used multi-document summarization corpora – Multinews and CQASumm, and observe an improvement of$+2.91$and$+6.64$ROUGE-L scores, respectively. The qualitative analyses on semantic similarity by BERTScore, faithfulness by question-answer evaluation and human evaluation show significant improvement over the baseline-generated summaries.

Due to the over-emphasize of the quantity of data, the data quality has often been overlooked. However, not all training data points contribute equally to learning. In particular, if mislabeled, it might actively damage the performance of the model and the ability to generalize out of distribution, as the model might end up learning spurious artifacts present in the dataset. This problem gets compounded by the prevalence of heavily parameterized and complex deep neural networks, which can, with their high capacity, end up memorizing the noise present in the dataset. This paper proposes a novel statistic - noise score, as a measure for the quality of each data point to identify such mislabeled samples based on the variations in the latent space representation. In our work, we use the representations derived by the inference network of data quality supervised variational autoencoder (AQUAVS). Our method leverages the fact that samples belonging to the same class will have similar latent representations. Therefore, by identifying the outliers in the latent space, we can find the mislabeled samples. We validate our proposed statistic through experimentation by corrupting MNIST, FashionMNIST, and CIFAR10/100 datasets in different noise settings for the task of identifying mislabelled samples. We further show significant improvements in accuracy for the classification task for each dataset.