Debiasing Algorithm through Model Adaptation

Large language models are becoming the go-to solution for various language tasks. However, with growing capacity, models are prone to rely on spurious correlations stemming from biases and stereotypes present in the training data. This work proposes a novel method for detecting and mitigating gender bias in language models. We perform causal analysis to identify problematic model components and discover that mid-upper feed-forward layers are most prone to convey biases. Based on the analysis results, we adapt the model by multiplying these layers by a linear projection. Our titular method, DAMA, significantly decreases bias as measured by diverse metrics while maintaining the model's performance on downstream tasks. We release code for our method and models, which retrain LLaMA's state-of-the-art performance while being significantly less biased.

SphereMap: Dynamic Multi-Layer Graph Structure for Rapid Safety-Aware UAV Planning

A flexible topological representation consisting of a two-layer graph structure built on-board an Unmanned Aerial Vehicle (UAV) by continuously filling the free space of an occupancy map with intersecting spheres is proposed in this \paper{}. Most state-of-the-art planning methods find the shortest paths while keeping the UAV at a pre-defined distance from obstacles. Planning over the proposed structure reaches this pre-defined distance only when necessary, maintaining a safer distance otherwise, while also being orders of magnitude faster than other state-of-the-art methods. Furthermore, we demonstrate how this graph representation can be converted into a lightweight shareable topological-volumetric map of the environment, which enables decentralized multi-robot cooperation. The proposed approach was successfully validated in several kilometers of real subterranean environments, such as caves, devastated industrial buildings, and in the harsh and complex setting of the final event of the DARPA SubT Challenge, which aims to mimic the conditions of real search and rescue missions as closely as possible, and where our approach achieved the \nth{2} place in the virtual track.

Independent Components of Word Embeddings Represent Semantic Features

Independent Component Analysis (ICA) is an algorithm originally developed for finding separate sources in a mixed signal, such as a recording of multiple people in the same room speaking at the same time. It has also been used to find linguistic features in distributional representations. In this paper, we used ICA to analyze words embeddings. We have found that ICA can be used to find semantic features of the words and these features can easily be combined to search for words that satisfy the combination. We show that only some of the independent components represent such features, but those that do are stable with regard to random initialization of the algorithm.

DialogueScript: Using Dialogue Agents to Produce a Script

We present a novel approach to generating scripts by using agents with different personality types. To manage character interaction in the script, we employ simulated dramatic networks. Automatic and human evaluation on multiple criteria shows that our approach outperforms a vanilla-GPT2-based baseline. We further introduce a new metric to evaluate dialogue consistency based on natural language inference and demonstrate its validity.

THEaiTRE 1.0: Interactive generation of theatre play scripts

We present the first version of a system for interactive generation of theatre play scripts. The system is based on a vanilla GPT-2 model with several adjustments, targeting specific issues we encountered in practice. We also list other issues we encountered but plan to only solve in a future version of the system. The presented system was used to generate a theatre play script planned for premiere in February 2021.

Measuring Memorization Effect in Word-Level Neural Networks Probing

Multiple studies have probed representations emerging in neural networks trained for end-to-end NLP tasks and examined what word-level linguistic information may be encoded in the representations. In classical probing, a classifier is trained on the representations to extract the target linguistic information. However, there is a threat of the classifier simply memorizing the linguistic labels for individual words, instead of extracting the linguistic abstractions from the representations, thus reporting false positive results. While considerable efforts have been made to minimize the memorization problem, the task of actually measuring the amount of memorization happening in the classifier has been understudied so far. In our work, we propose a simple general method for measuring the memorization effect, based on a symmetric selection of comparable sets of test words seen versus unseen in training. Our method can be used to explicitly quantify the amount of memorization happening in a probing setup, so that an adequate setup can be chosen and the results of the probing can be interpreted with a reliability estimate. We exemplify this by showcasing our method on a case study of probing for part of speech in a trained neural machine translation encoder.

THEaiTRE: Artificial Intelligence to Write a Theatre Play

We present THEaiTRE, a starting project aimed at automatic generation of theatre play scripts. This paper reviews related work and drafts an approach we intend to follow. We plan to adopt generative neural language models and hierarchical generation approaches, supported by summarization and machine translation methods, and complemented with a human-in-the-loop approach.

Semantic Holism and Word Representations in Artificial Neural Networks

Artificial neural networks are a state-of-the-art solution for many problems in natural language processing. What can we learn about language and meaning from the way artificial neural networks represent it? Word representations obtained from the Skip-gram variant of the word2vec model exhibit interesting semantic properties. This is usually explained by referring to the general distributional hypothesis, which states that the meaning of the word is given by the contexts where it occurs. We propose a more specific approach based on Frege's holistic and functional approach to meaning. Taking Tugendhat's formal reinterpretation of Frege's work as a starting point, we demonstrate that it is analogical to the process of training the Skip-gram model and offers a possible explanation of its semantic properties.

A Test Suite and Manual Evaluation of Document-Level NMT at WMT19

As the quality of machine translation rises and neural machine translation (NMT) is moving from sentence to document level translations, it is becoming increasingly difficult to evaluate the output of translation systems. We provide a test suite for WMT19 aimed at assessing discourse phenomena of MT systems participating in the News Translation Task. We have manually checked the outputs and identified types of translation errors that are relevant to document-level translation.

Derivational Morphological Relations in Word Embeddings

Derivation is a type of a word-formation process which creates new words from existing ones by adding, changing or deleting affixes. In this paper, we explore the potential of word embeddings to identify properties of word derivations in the morphologically rich Czech language. We extract derivational relations between pairs of words from DeriNet, a Czech lexical network, which organizes almost one million Czech lemmata into derivational trees. For each such pair, we compute the difference of the embeddings of the two words, and perform unsupervised clustering of the resulting vectors. Our results show that these clusters largely match manually annotated semantic categories of the derivational relations (e.g. the relation 'bake--baker' belongs to category 'actor', and a correct clustering puts it into the same cluster as 'govern--governor').