Test of Time Awards

Presidential Address

One of the frequent points in the mainstream narrative about large language models is that they have “emergent properties” (sometimes even dangerous enough to be considered existential risk to mankind). However, there is much disagreement about even the very definition of such properties. If they are understood as a kind of generalization beyond training data - as something that a model does without being explicitly trained for it - I argue that we have not in fact established the existence of any such properties, and at the moment we do not even have the methodology for doing so.

A sanity check on emergent properties

Transformer-based models have achieved dominant performance in numerous NLP tasks. Despite their remarkable successes, pre-trained transformers such as BERT suffer from a computationally expensive self-attention mechanism that interacts with all tokens, including the ones unfavorable to classification performance. To overcome these challenges, we propose integrating two strategies: token pruning and token combining. Token pruning eliminates less important tokens in the attention mechanism’s key and value as they pass through the layers. Additionally, we adopt fuzzy logic to handle uncertainty and alleviate potential mispruning risks arising from an imbalanced distribution of each token’s importance. Token combining, on the other hand, condenses input sequences into smaller sizes in order to further compress the model. By integrating these two approaches, we not only improve the model’s performance but also reduce its computational demands. Experiments with various datasets demonstrate superior performance compared to baseline models, especially with the best improvement over the existing BERT model, achieving +5%p in accuracy and +5.6%p in F1 score. Additionally, memory cost is reduced to 0.61x, and a speedup of 1.64x is achieved.

Focus on the Core: Efficient Attention via Pruned Token Compression for Document Classification

Emotion arcs capture how an individual (or a population) feels over time. They are widely used in industry and research; however, there is little work on evaluating the automatically generated arcs. This is because of the difficulty of establishing the true (gold) emotion arc. Our work, for the first time, systematically and quantitatively evaluates automatically generated emotion arcs. We also compare two common ways of generating emotion arcs: Machine-Learning (ML) models and Lexicon-Only (LexO) methods. By running experiments on 18 diverse datasets in 9 languages, we show that despite being markedly poor at instance level emotion classification, LexO methods are highly accurate at generating emotion arcs when aggregating information from hundreds of instances. We also show, through experiments on six indigenous African languages, as well as Arabic, and Spanish, that automatic translations of English emotion lexicons can be used to generate high-quality emotion arcs in less-resource languages. This opens up avenues for work on emotions in languages from around the world; which is crucial for commerce, public policy, and health research in service of speakers often left behind. Code and resources: https://github.com/dteodore/EmotionArcs

Evaluating Emotion Arcs Across Languages: Bridging the Global Divide in Sentiment Analysis

Many real-world tasks involve a mixed-initiative setup, wherein humans and AI systems collaboratively perform a task. While significant work has been conducted towards enabling humans to specify, through language, exactly how an agent should complete a task (i.e., low-level specification), prior work lacks on interpreting the high-level strategic intent of the human commanders. Parsing strategic intent from language will allow autonomous systems to independently operate according to the user's plan without frequent guidance or instruction. In this paper, we build a computational interface capable of translating unstructured language strategies into actionable intent in the form of goals and constraints. Leveraging a game environment, we collect a dataset of over 1000 examples, mapping language strategies to the corresponding goals and constraints, and show that our model, trained on this dataset, significantly outperforms human interpreters in inferring strategic intent (i.e., goals and constraints) from language (p < 0.05). Furthermore, we show that our model (125M parameters) significantly outperforms ChatGPT for this task (p < 0.05) in a low-data setting.

A Computational Interface to Translate Strategic Intent from Unstructured Language in a Low-Data Setting | VIDEO

Identifying intents from dialogue utterances forms an integral component of task-oriented dialogue systems. Intent-related tasks are typically formulated either as a classification task, where the utterances are classified into predefined categories or as a clustering task when new and previously unknown intent categories need to be discovered from these utterances. Further, the intent classification may be modeled in a multiclass (MC) or multilabel (ML) setup. While typically these tasks are modeled as separate tasks, we propose INTENDD a unified approach leveraging a shared utterance encoding backbone. INTENDD uses an entirely unsupervised contrastive learning strategy for representation learning, where pseudo labels for the unlabeled utterances are generated based on their lexical features. Additionally, we introduce a two-step post-processing setup for the classification tasks using modified adsorption. Here, first, the residuals in the training data are propagated followed by smoothing the labels both modeled in a transductive setting. Through extensive evaluations on various benchmark datasets, we find that our approach consistently outperforms competitive baselines across all three tasks. On average, INTENDD reports percentage improvements of 2.32 %, 1.26 %, and 1.52 % in their respective metrics for few-shot MC, few-shot ML, and the intent discovery tasks respectively.

IntenDD: A Unified Contrastive Learning Approach for Intent Detection and Discovery

Open-domain dialogue system usually requires different sources of knowledge to generate more informative and evidential responses. However, existing knowledge-grounded dialogue systems either focus on a single knowledge source or overlook the dependency between multiple sources of knowledge, which may result in generating inconsistent or even paradoxical responses. To incorporate multiple knowledge sources and dependencies between them, we propose SAFARI, a novel framework that leverages the exceptional capabilities of large language models (LLMs) in planning, understanding, and incorporating under both supervised and unsupervised settings. Specifically, SAFARI decouples the knowledge grounding into multiple sources and response generation, which allows easy extension to various knowledge sources including the possibility of not using any sources. To study the problem, we construct a personalized knowledge-grounded dialogue dataset Knowledge Behind Persona (KBP), which is the first to consider the dependency between persona and implicit knowledge. Experimental results on the KBP dataset demonstrate that the SAFARI framework can effectively produce persona-consistent and knowledge-enhanced responses.

Large Language Models as Source Planner for Personalized Knowledge-grounded Dialogues

Ontonotes has served as the most important benchmark for coreference resolution. However, for ease of annotation, several long documents in Ontonotes were split into smaller parts. In this work, we build a corpus of coreference-annotated documents of significantly longer length than what is currently available. We do so by providing an accurate, manually-curated, merging of annotations from documents that were split into multiple parts in the original Ontonotes annotation process. The resulting corpus, which we call LongtoNotes contains documents in multiple genres of the English language with varying lengths, the longest of which are up to 8x the length of documents in Ontonotes, and 2x those in Litbank. We evaluate state-of-the-art neural coreference systems on this new corpus, analyze the relationships between model architectures/hyperparameters and document length on performance and efficiency of the models, and demonstrate areas of improvement in long-document coreference modeling revealed by our new corpus.

LongtoNotes: OntoNotes with Longer Coreference Chain

Inverse Text Normalization (ITN) involves rewriting the verbalised form of text from spoken transcripts to its corresponding written form. The task inherently expects challenges in identifying ITN entries due to spelling variations in words arising out of dialects, transcription errors etc. Additionally, in Tamil, word boundaries between adjacent words in a sentence often get obscured due to Punarchi, i.e. phonetic transformation of these boundaries. Being morphologically rich, the words in Tamil show a high degree of agglutination due to inflection and clitics. The combination of such factors leads to a high degree of surfaceform variations, making scalability with pure rule-based approaches difficult. Instead, we experiment with fine-tuning three pre-trained neural LMs, consisting of a seq2seq model (s2s), a non-autoregressive text editor (NAR) and a sequence tagger + rules combination (tagger). While the tagger approach works best in a fully-supervised setting, s2s performs the best (98.05 F-Score) when augmented with additional data, via bootstrapping and data augmentation (DA&B). S2S reports a cumulative percentage improvement of 20.1 %, and statistically significant gains for all our models with DA&B. Compared to a fully supervised setup, bootstrapping alone reports a percentage improvement as high as 14.12 %, even with a small seed set of 324 ITN entries.

Scaling Neural ITN for Numbers and Temporal Expressions in Tamil: Findings for an Agglutinative Low-resource Language | VIDEO

A key contribution to being a successful researcher in natural language processing, as in any area, is having a clear overarching vision of what your body of research is trying to accomplish. Using my own 40-year career as an example, I will attempt to provide general advice on formulating and pursuing a coherent research vision. In particular, I will focus on formulating a unique, personal objective that exploits your specific talents, knowledge, and passions, and that is distinct from the current popular trends in the field. I will also focus on formulating a vision that bridges existing fields of study to produce an overarching agenda that unifies previously disparate ideas.

The Vision Thing: Finding and Pursuing your Research Passion

Large language models (LLMs) have shown increasing capacity at planning and executing a high-level goal in a live computer environment (e.g. MiniWoB++). To perform a task, recent works often require a model to learn from trace examples of the task via either supervised learning or few/many-shot prompting. Without these trace examples, it remains a challenge how an agent can autonomously learn and improve its control on a computer, which limits the ability of an agent to perform a new task. We approach this problem with a zero-shot agent that requires no given expert traces. Our agent plans for executable actions on a partially observed environment, and iteratively progresses a task by identifying and learning from its mistakes via self-reflection and structured thought management. On the easy tasks of MiniWoB++, we show that our zero-shot agent often outperforms recent SoTAs, with more efficient reasoning. For tasks with more complexity, our reflective agent performs on par with prior best models, even though previous works had the advantages of accessing expert traces or additional screen information.

Underline digital video library