Thailand

The rapid growth in natural language processing (NLP) research has led to numerous new models, outpacing our understanding of how they compare to established ones. One major reason for this difficulty is saturating benchmarks, which may not well reflect differences in model performance in the wild. In this work, we introduce a novel framework to compare two NLP models by revealing their shared invariance to interpretable input perturbations targeting a specific linguistic capability. Via experiments on models from the same and different architecture families, this framework offers insights about how changes in models (e.g., distillation, size increase) affect linguistic capabilities. Furthermore, our framework enables evaluation of invariances between commercial black-box models (e.g., InstructGPT family) and models that are better understood (e.g., GPT-2). Across experiments, we observe that large language models share many invariances encoded by models of various sizes, whereas the invariances by large models are only shared by other large models. Possessing a wide variety of invariances may be key to the recent successes of large language models, and our framework can shed light on the types of invariances retained or emerging in new models. We make the code publicly available.

ACL 2024

Perturbed examples reveal invariances shared by language models

shared invariance

linguistic capabilities

behavioral comparison

poster

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.

You are required to register for this event. **Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.

Please register!

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

Explainable AI methods facilitate the understanding of model behaviour, yet, small, imperceptible perturbations to inputs can vastly distort explanations. As these explanations are typically evaluated holistically, before model deployment, it is difficult to assess when a particular explanation is trustworthy. Some studies have tried to create confidence estimators for explanations, but none have investigated an existing link between uncertainty and explanation quality. We artificially simulate epistemic uncertainty in text input by introducing noise at inference time. In this large-scale empirical study, we insert different levels of noise perturbations and measure the effect on the output of pre-trained language models and different uncertainty metrics. Realistic perturbations have minimal effect on performance and explanations, yet masking has a drastic effect. We find that high uncertainty doesn't necessarily imply low explanation plausibility; the correlation between the two metrics can be moderately positive when noise is exposed during the training process. This suggests that noise-augmented models may be better at identifying salient tokens when uncertain. Furthermore, when predictive and epistemic uncertainty measures are over-confident, the robustness of a saliency map to perturbation can indicate model stability issues. Integrated Gradients shows the overall greatest robustness to perturbation, while still showing model-specific patterns in performance; however, this phenomenon is limited to smaller Transformer-based language models.

Investigating the Impact of Model Instability on Explanations and Uncertainty

The Spoken Language Understanding Evaluation (SLUE) suite of benchmark tasks was recently introduced to address the need for open
resources and benchmarking of complex spoken language understanding (SLU) tasks, including both classification and sequence generation tasks, on natural speech. The benchmark has demonstrated preliminary success in using pre-trained speech foundation models (SFM) for these SLU tasks. However, the community still lacks a fine-grained understanding of the comparative utility of different SFMs. Inspired by this, we ask: which SFMs offer the most benefits for these complex SLU tasks, and what is the most effective approach for incorporating these SFMs? To answer this, we perform an extensive evaluation of multiple supervised and self-supervised SFMs using several evaluation protocols: (i) frozen SFMs with a lightweight prediction head, (ii) frozen SFMs with a complex prediction head, and (iii) fine-tuned SFMs with a lightweight prediction head. Although the supervised SFMs are pre-trained on much more speech recognition data (with labels), they do not always outperform self-supervised SFMs; the latter tend to perform at least as well as, and sometimes better than, supervised SFMs, especially on the sequence generation tasks in SLUE. While there is no universally optimal way of incorporating SFMs, the complex prediction head gives the best performance for most tasks, although it increases the inference time. We also introduce an open-source toolkit and performance leaderboard, SLUE-PERB, for these tasks and modeling strategies.

On the Evaluation of Speech Foundation Models for Spoken Language Understanding

LLMs have demonstrated remarkable capability for understanding semantics, but their understanding of pragmatics is not well studied. To this end, we release a Pragmatics Understanding Benchmark (PUB) dataset consisting of fourteen tasks in four pragmatics phenomena, namely; Implicature, Presupposition, Reference, and Deixis. We curate high-quality test sets for each task, consisting of Multiple Choice Question Answers (MCQA). PUB includes a total of 28k data points, 6.1k are newly annotated. We evaluate nine models varying in the number of parameters and type of training. Our study reveals several key observations about the pragmatic capabilities of LLMs: 1. chat-fine-tuning strongly benefits smaller models, 2. large base models are competitive with their chat-fine-tuned counterparts, 3. there is a huge variance in performance across different pragmatics phenomena, and 4. a noticeable performance gap between human capabilities and model capabilities. We hope that PUB will enable comprehensive evaluation of LLM's pragmatic reasoning capabilities.

PUB: A Pragmatics Understanding Benchmark for Assessing LLMs' Pragmatics Capabilities

This paper introduces EmoTransKG, an innovative Emotion Knowledge Graph (EKG) that establishes connections and transformations between emotions across diverse open-textual events. Compared to existing EKGs, which primarily focus on linking emotion keywords to related terms or on assigning sentiment dimension ratings to emotion words by humans, EmoTransKG aims to represent the general knowledge involved in emotion transformation. Specifically, in conversations, successive emotions expressed by a single speaker are temporally considered as the head and tail entities, with open-text utterances (events) occurring between them representing the relation. To explore the knowledge of emotion transformations described in EmoTransKG, we develop a Transformer-based translational model called EmoTransNet, which predictively trains tail entities by interpreting the relation as an operation that transforms the source emotion into the target emotion. Particularly, our designed EmoTransNet serves as a plug-in module that seamlessly integrates with any conversational emotion recognition (CER) models for emotion retrofitting. Experimental results on two CER datasets demonstrate that the incorporation of EmoTransNet with baseline models results in substantial improvements, and the qualitative visualization of entities and relations clearly clarify their unique roles in emotion transformations. These experiments confirm the quality and effectiveness of EmoTransKG.

EmoTransKG: An Innovative Emotion Knowledge Graph to Reveal Emotion Transformation

While supervised fine-tuning (SFT) has been a straightforward approach for tailoring the output of foundation large language model (LLM) to specific preferences, concerns have been raised about the depth of this alignment, with some critiques suggesting it is merely “superficial”. We critically examine this hypothesis within the scope of cross-lingual generation tasks, proposing that the effectiveness of SFT may be constrained by its reliance on prior tokens to guide cross-lingual generation. Based on this crucial insight, and in response to the challenges posed by the costly and limited availability of non-English data for SFT, we introduce a novel training-free alignment method named PreTTY, which employs minimal task-related prior tokens to bridge the foundation LLM and the SFT LLM, achieving comparable performance without training. Experiments on machine translation and part-of-speech tagging across seven languages demonstrate the efficacy of PreTTY in cross-lingual settings. Remarkably, by initiating the decoding process with only one or two prior tokens, foundation LLMs can attain up to 98\% of the performance metrics of their SFT counterparts. This method presents a cost-effective alternative to traditional SFT and advances the democratization of multilingual LLMs.

Prefix Text as a Yarn: Eliciting Non-English Alignment in Foundation Language Model

Multilingual Large Language Models (LLMs) achieve remarkable levels of zero-shot cross-lingual transfer performance. We speculate that this is predicated on their ability to align languages without explicit supervision from parallel sentences. While representations of translationally equivalent sentences in different languages are known to be similar after convergence, however, it remains unclear how such cross-lingual alignment emerges during pre-training of LLMs. Our study leverages intrinsic probing techniques, which identify which subsets of neurons encode linguistic features, to correlate the degree of cross-lingual neuron overlap with the zero-shot cross-lingual transfer performance for a given model. In particular, we rely on checkpoints of BLOOM, a multilingual autoregressive LLM, across different training steps and model scales. We observe a high correlation between neuron overlap and downstream performance, which supports our hypothesis on the conditions leading to effective cross-lingual transfer. Interestingly, we also detect a degradation of both implicit alignment and multilingual abilities in certain phases of the pre-training process, providing new insights into the multilingual pretraining dynamics.

Probing the Emergence of Cross-lingual Alignment during LLM Training

In the past year, MultiModal Large Language Models (MM-LLMs) have undergone substantial advancements, augmenting off-the-shelf LLMs to support MM inputs or outputs via cost-effective training strategies. The resulting models not only preserve the inherent reasoning and decision-making capabilities of LLMs but also empower a diverse range of MM tasks. In this paper, we provide a comprehensive survey aimed at facilitating further research of MM-LLMs. Initially, we outline general design formulations for model architecture and training pipeline. Subsequently, we introduce a taxonomy encompassing $126$ MM-LLMs, each characterized by its specific formulations. Furthermore, we review the performance of selected MM-LLMs on mainstream benchmarks and summarize key training recipes to enhance the potency of MM-LLMs. Finally, we explore promising directions for MM-LLMs while concurrently maintaining a [real-time tracking website](https://mm-llms.github.io/) for the latest developments in the field. We hope that this survey contributes to the ongoing advancement of the MM-LLMs domain.

MM-LLMs: Recent Advances in MultiModal Large Language Models

Understanding and following natural language instructions while navigating through complex, real-world environments poses a significant challenge for general-purpose robots. These environments often include obstacles and pedestrians, making it essential for autonomous agents to possess the capability of self-corrected planning to adjust their actions based on feedback from the surroundings. However, the majority of existing vision-and-language navigation (VLN) methods primarily operate in less realistic simulator settings and do not incorporate environmental feedback into their decision-making processes. To address this gap, we introduce a novel zero-shot framework called CorNav, utilizing a large language model for decision-making and comprising two key components: 1) incorporating environmental feedback for refining future plans and adjusting its actions, and 2) multiple domain experts for parsing instructions, scene understanding, and refining predicted actions. In addition to the framework, we develop a 3D simulator that renders realistic scenarios using Unreal Engine 5. To evaluate the effectiveness and generalization of navigation agents in a zero-shot multi-task setting, we create a benchmark called NavBench. Our empirical study involves deploying 7 baselines across four tasks, i.e., goal-conditioned navigation given a specific object category, goal-conditioned navigation given simple instructions, finding abstract objects based on high-level instructions, and step-by-step instruction following. Extensive experiments demonstrate that CorNav consistently outperforms all baselines by a significant margin across all tasks. On average, CorNav achieves a success rate of 28.1%, surpassing the best baseline's performance of 20.5%.

CorNav: Autonomous Agent with Self-Corrected Planning for Zero-Shot Vision-and-Language Navigation

Whataboutism, a potent tool for disrupting narratives and sowing distrust, remains under-explored in quantitative NLP research. Moreover, past work has not distinguished its use as a strategy for misinformation and propaganda from its use as a tool for pragmatic and semantic framing. We introduce new datasets from Twitter/X and YouTube, revealing overlaps as well as distinctions between whataboutism, propaganda, and the tu quoque fallacy. Furthermore, drawing on recent work in linguistic semantics, we differentiate the `what about' lexical construct from whataboutism. Our experiments bring to light unique challenges in its accurate detection, prompting the introduction of a novel method using attention weights for negative sample mining. We report significant improvements of 4% and 10% over previous state-of-the-art methods in our Twitter and YouTube collections, respectively.

Paying Attention to Deflections: Mining Pragmatic Nuances for Whataboutism Detection in Online Discourse

Despite great performance on many tasks, language models (LMs) still struggle with reasoning, sometimes providing responses that cannot possibly be true because they stem from logical incoherence. We call such responses strong hallucinations and prove that they follow from an LM's computation of its internal representations for logical operators and outputs from those representations. Focusing on negation, we provide a novel solution in which negation is treated not as another element of a latent representation, but as an operation over an LM's latent representations that constrains how they may evolve}. We show that our approach improves model performance in cloze prompting and natural language inference tasks with negation without requiring training on sparse negative data.

Premium content

Perturbed examples reveal invariances shared by language models

Downloads

Next from ACL 2024

Investigating the Impact of Model Instability on Explanations and Uncertainty

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES