Ireland

Pre-trained models (PTMs) have lead to great improvements in natural language generation (NLG). However, it is still unclear how much commonsense knowledge they possess. With the goal of evaluating commonsense knowledge of NLG models, recent work has proposed the problem of generative commonsense reasoning, e.g., to compose a logical sentence given a set of unordered concepts. Existing approaches to this problem hypothesize that PTMs lack sufficient parametric knowledge for this task, which can be overcome by introducing external knowledge or task-specific pre-training objectives. Different from this trend, we argue that PTM&#39;s \textit{inherent} ability for generative commonsense reasoning is underestimated due to the order-agnostic property of its input. In particular, we hypothesize that the order of the input concepts can affect the PTM&#39;s ability to utilize its commonsense knowledge. To this end, we propose a pre-ordering approach to elaborately manipulate the order of the given concepts before generation. Experiments show that our approach can outperform the more sophisticated models that have access to a lot of external data and resources.

ACL 2022

Revisiting Generative Commonsense Reasoning: A Pre-Ordering Approach

pre-ordering

commongen

generative commonsense reasoning

# Welcome everyone to ACL 2022!

The 60th Annual Meeting of the Association for Computational Linguistics is taking place May 22-27, 2022 as a hybrid event, in Dublin and online. We are happy to welcome all of you to this anniversary edition with an almost 50-50 in-person and virtual participation. 
The main conference program features oral presentations, in-person and virtual posters and demo sessions, a plenary session for our best paper presentations and awards, three amazing keynote events and two new initiatives of invited talks: Spotlight Talks for Young Rising Stars (STIRS) and The Next Big Idea Talks. Posters (including Findings of ACL 2022) and demos are grouped by areas for both the in-person and the virtual sessions. For the virtual component, the talks will be on Zoom and the posters and the demos will be in GatherTown. The Student Research Workshop will have an oral session and a poster session as part of Poster Session 1. The program also features eight Tutorials and 28 Workshops. 

 
We wish you a wonderful conference! 
[**The ACL 2022 Organizing Committee**](https://www.2022.aclweb.org/organisers)
 
[**Conference Handbook**](https://drive.google.com/file/d/1_BUCMfhMVrjG9E2e71aHdHeE28KSje0l/view?usp=sharing) 
[**Mini Handbook**](https://drive.google.com/file/d/1qlBKl0wzmlVF1oCeMQl3BahLd9nLP5Ce/view?usp=sharing) 
[**Posters and Demo guides**](https://drive.google.com/file/d/1UucMAoCNncIOaH1rMMDa0owuG9qgvJTG/view?usp=sharing)

The Association for Computational Linguistics (ACL) is the premier international scientific and professional society for people working on computational problems involving human language, a field often referred to as either computational linguistics or natural language processing (NLP). 

Pre-trained models (PTMs) have lead to great improvements in natural language generation (NLG). However, it is still unclear how much commonsense knowledge they possess. With the goal of evaluating commonsense knowledge of NLG models, recent work has proposed the problem of generative commonsense reasoning, e.g., to compose a logical sentence given a set of unordered concepts. Existing approaches to this problem hypothesize that PTMs lack sufficient parametric knowledge for this task, which can be overcome by introducing external knowledge or task-specific pre-training objectives. Different from this trend, we argue that PTM's \textit{inherent} ability for generative commonsense reasoning is underestimated due to the order-agnostic property of its input. In particular, we hypothesize that the order of the input concepts can affect the PTM's ability to utilize its commonsense knowledge. To this end, we propose a pre-ordering approach to elaborately manipulate the order of the given concepts before generation. Experiments show that our approach can outperform the more sophisticated models that have access to a lot of external data and resources.

workshop paper

Pre-trained language models have shown successful progress in many text understanding benchmarks. This work explores the capability of these models to predict actionable plans in real-world environments. Given a text instruction, we show that language priors encoded in pre-trained models allow us to infer fine-grained subgoal sequences. In contrast to recent methods which make strong assumptions about subgoal supervision, our experiments show that language models can infer detailed subgoal sequences from few training sequences without any fine-tuning. We further propose a simple strategy to re-rank language model predictions based on interaction and feedback from the environment.Combined with pre-trained navigation and visual reasoning components, our approach demonstrates competitive performance on subgoal prediction and task completion in the ALFRED benchmark compared to prior methods that assume more subgoal supervision.

Few-shot Subgoal Planning with Language Models

Commonsense is defined as the knowledge that is shared by everyone. However, certain types of commonsense knowledge are correlated with culture and geographic locations and they are only shared locally. For example, the scenarios of wedding ceremonies vary across regions due to different customs influenced by historical and religious factors. Such regional characteristics, however, are generally omitted in prior work. In this paper, we construct a Geo-Diverse Visual Commonsense Reasoning dataset (GD-VCR) to test vision-and-language models' ability to understand cultural and geo-location-specific commonsense. In particular, we study two state-of-the-art Vision-and-Language models, VisualBERT and ViLBERT trained on VCR, a standard multimodal commonsense benchmark with images primarily from Western regions. We then evaluate how well the trained models can generalize to answering the questions in GD-VCR. We find that the performance of both models for non-Western regions including East Asia, South Asia, and Africa is significantly lower than that for Western region. We analyze the reasons behind the performance disparity and find that the performance gap is larger on QA pairs that: 1) are concerned with culture-related scenarios, e.g., weddings, religious activities, and festivals; 2) require high-level geo-diverse commonsense reasoning rather than low-order perception and recognition.

Broaden the Vision: Geo-Diverse Visual Commonsense Reasoning

Specialized number representations in NLP have shown improvements on numerical reasoning tasks like arithmetic word problems and masked number prediction. But humans also use numeracy to make better sense of world concepts, e.g., you can seat 5 people in your ‘room’ but not 500. Does a better grasp of numbers improve a model’s understanding of other concepts and words? This paper studies the effect of using six different number encoders on the task of masked word prediction (MWP), as a proxy for evaluating literacy. To support this investigation, we develop Wiki-Convert, a 900,000 sentence dataset annotated with numbers and units, to avoid conflating nominal and ordinal number occurrences. We find a significant improvement in MWP for sentences containing numbers, that exponent embeddings are the best number encoders, yielding over 2 points jump in prediction accuracy over a BERT baseline, and that these enhanced literacy skills also generalize to contexts without annotated numbers. We release all code at https://git.io/JuZXn.

Numeracy enhances the Literacy of Language Models

Robots operating in everyday environments need to effectively perceive, model, and infer semantic properties of objects. Existing knowledge reasoning frameworks only model binary relations between an object's class label and its semantic properties, unable to collectively reason about object properties detected by different perception algorithms and grounded in diverse sensory modalities. We bridge the gap between multimodal perception and knowledge reasoning by introducing an n-ary representation that models complex, inter-related object properties. To tackle the problem of collecting n-ary semantic knowledge at scale, we propose a transformer neural network that directly generalizes knowledge from observations of object instances. The learned model can reason at different levels of abstraction, effectively predicting unknown properties of objects in different environmental contexts given different amounts of observed information. We quantitatively validate our approach against five prior methods on LINK, a unique dataset we contribute that contains 1457 situated object instances with 15 multimodal properties types and 200 total properties. Additionally, we apply our work to a mobile manipulation robot, demonstrating its ability to leverage n-ary reasoning to retrieve objects and actively detect object properties.

Learning Instance-Level N-Ary Semantic Knowledge At Scale For Robots Operating in Everyday Environments

The ability to recognize analogies is fundamental to human cognition. Existing benchmarks to test word analogy do not reveal the underneath process of analogical reasoning of neural models. Holding the belief that models capable of reasoning should be right for the right reasons, we propose a first-of-its-kind Explainable Knowledge-intensive Analogical Reasoning benchmark (E-KAR). Our benchmark consists of 1,655 (in Chinese) and 1,251 (in English) problems sourced from the Civil Service Exams, which require intensive background knowledge to solve. More importantly, we design a free-text explanation scheme to explain whether an analogy should be drawn, and manually annotate them for each and every question and candidate answer. Empirical results suggest that this benchmark is very challenging for some state-of-the-art models for both explanation generation and analogical question answering tasks, which invites further research in this area.

E-KAR: A Benchmark for Rationalizing Natural Language Analogical Reasoning

Answering questions about why events happen in narratives requires commonsense knowledge that is external to the narrative. What aspects of this knowledge is accessible to large models? What aspects can be made accessible via exter- nal commonsense resources? We study these in the context of answering Why questions in the TellMeWhy dataset using COMET as a source of relevant commonsense relations. We analyze the relative improvements over a base T5 model when (a) increasing the model size, (b) injecting knowledge from COMET as part of the task input, and (c) asking the model to generate COMET relation type as an explanation in addition to its answer. Results show that the larger model, as expected, yields substantial improvements over the base. Interestingly, we find that the question specific COMET relations can provide substantial improvements for both base and large models, with additional possible gains when asking the model to also generate COMET relation type. So, we augment a large model with noisy hints from COMET and find that this improves performance on the TellMeWhy task. We also develop a simple ontology of knowledge types and analyze the relative cover- age of the different models on these categories. Together, these findings suggest potential for methods that can automatically select and inject commonsense from relevant sources.

Analyzing the Contribution of Commonsense Knowledge for Why-Question Answering

Large LMs such as GPT-3 are powerful, but can commit mistakes that are obvious to hu- mans. For example, GPT-3 would mistakenly interpret "What word is similar to good?" to mean a homonym, while the user intended a synonym. Our goal is to effectively correct such errors via user interactions with the system but without retraining, which will be prohibitively costly. We pair GPT-3 with a growing memory of recorded cases where the model misunderstood the user’s intents, along with user feedback for clarification. Such a memory allows our system to produce enhanced prompts for any new query based on the user feedback for error correction on similar cases in the past. On four tasks (two lexical tasks, two advanced ethical reasoning tasks), we show how a (simulated) user can interactively teach a deployed GPT-3, substantially in- creasing its accuracy over the queries with different kinds of misunderstandings by the GPT- 3. Our approach is a step towards the low-cost utility enhancement for very large pre-trained LMs.

Memory-assisted prompt editing to improve GPT-3 after deployment

How grounded is document-grounded conversational AI?

Artificial Intelligence services are increasingly being used to provide accessibility for
consumers with sensory disabilities in information and communication technologies.
However, these AI-powered are only as good as the data they’ve been fed, which
often reflects historical or current biases either in the collection of data or in the data
itself. In general, people with diverse visual and auditory disabilities are not wellrepresented in these datasets because although many people have some disability,
the number with any one disability is usually a small percentage of the population.
These disparities in representation can lead to disproportionate impacts of AIpowered services on people with disabilities, and though laws to protect people with
disabilities exist, the evaluation of AI-based systems and the models they rely on, to
comply with those laws in both the letter and spirit of the law are important, unsolved
problems. We advocate for guidelines for an evaluation of the models used by AIpowered services that considers their fairness, accountability, transparency, in terms
of use-cases, inclusiveness, and legal requirements for accessibility. These
accessibility guidelines can be used by architects or developers of AI-powered
services to ensure the services and underlying models are fair by being available to
consumers with diverse sensory abilities, accountable by complying with legal
requirements, and transparent by complying with both social and user expectations.

Accessibility Guidelines for AI-powered Services and Models

Humans quite frequently interact with conversational agents. The rapid advancement in generative language modeling through neural networks has helped advance the creation of intelligent conversational agents. Researchers typically evaluate the output of their models through crowdsourced judgments, but there are no established best practices for conducting such studies. We look closely at the practices of evaluation of NLG output, and discuss implications of human cognitive biases on experiment design and the resulting data.

Downloads

Next from ACL 2022

Few-shot Subgoal Planning with Language Models

Similar lecture

Diversifying Content Generation for Commonsense Reasoning with Mixture of Knowledge Graph Experts