Over the past two decades, numerous studies have demonstrated how less predictable (i.e. higher surprisal) words take more time to read. In general, these previous studies implicitly assumed the reading process to be purely responsive: readers observe a new word and allocate time to read it as required. These results, however, are also compatible with a reading time that is anticipatory: readers could, e.g., allocate time to a future word based on their expectation about it. In this work, we examine the anticipatory nature of reading by looking at how people's predictions about upcoming material influence reading times. Specifically, we test anticipation by looking at the effects of surprisal and contextual entropy on four reading-time datasets: two self-paced and two eye-tracking. In three of four datasets tested, we find that the entropy predicts reading times as well as (or better than) the surprisal. We then hypothesise four cognitive mechanisms through which the contextual entropy could impact RTs -- three of which we design experiments to analyse. Overall, our results support a view of reading that is both anticipatory and responsive.
Despite significant progress in the quality of language generated from abstractive summarization models, these models still exhibit the tendency to hallucinate, i.e., output content not supported by the source document. A number of works have tried to fix--or at least uncover the source of--the problem with limited success. In this paper, we identify a simple criterion under which models are significantly more likely to assign more probability to hallucinated content during generation: high model uncertainty. This finding offers a potential explanation for hallucinations: models default to favoring text with high marginal probability, i.e., high-frequency occurrences in the training set, when uncertain about a continuation. It also motivates possible routes for real-time intervention during decoding to prevent such hallucinations. We propose a decoding strategy that switches to optimizing for pointwise mutual information of the source and target token--rather than purely the probability of the target token--when the model exhibits uncertainty. Experiments on the XSum dataset show that our method decreases the probability of hallucinated tokens while maintaining the Rouge and BertS scores of top-performing decoding strategies.
While probabilistic language generators have improved dramatically over the last few years, the automatic evaluation metrics used to assess them have not kept pace with this progress. In the domain of language generation, a good metric must correlate highly with human judgements. Yet, with few exceptions, there is a lack of such metrics in the literature. In this work, we analyse the general paradigm of language generator evaluation. We first discuss the computational and qualitative issues with using automatic evaluation metrics that operate on probability distributions over strings, the backbone of most language generators. We then propose the use of distributions over clusters instead, where we cluster strings based on their text embeddings (obtained from a pretrained language model). While we find the biases introduced by this substitution to be quite strong, we observe that, empirically, this methodology leads to metric estimators with higher correlation with human judgements, while simultaneously reducing estimator variance. We finish the paper with a probing analysis, which leads us to conclude that -- by encoding syntactic- and coherence-level features of text, while ignoring surface-level features -- these clusters may simply be better equipped to evaluate state-of-the-art language models.
Probing has become a go-to methodology for interpreting and analyzing deep neural models in natural language processing. Yet recently, there has been much debate around the limitations and weaknesses of probes. In this work, we suggest a naturalistic strategy for input-level intervention on real world data in Spanish, which is a language with gender marking. Using our approach, we isolate morpho-syntactic features from counfounders in sentences, e.g. topic, which will then allow us to causally probe pre-trained models. We apply this methodology to analyze causal effects of gender and number on contextualized representations extracted from pre-trained models -- BERT, RoBERTa and GPT-2. Our experiments suggest that naturalistic intervention can give us stable estimates of causal effects, which varies across different words in a sentence. We further show the utility of our estimator in investigating gender bias in adjectives, and answering counterfactual questions in masked prediction. Our probing experiments highlights the importance of conducting causal probing in determining if a particular property is encoded in representations.
Shannon entropy is often a quantity of interest to linguists studying the communicative capacity of human language. However, entropy must typically be estimated from observed data because researchers do not have access to the underlying probability distribution that gives rise to these data. While entropy estimation is a well-studied problem in other fields, there is not yet a comprehensive exploration of the efficacy of entropy estimators for use with linguistic data. In this work, we fill this void, studying the empirical effectiveness of different entropy estimators for linguistic distributions. In a replication of two recent information-theoretic linguistic studies, we find evidence that the reported effect size is over-estimated due to over-reliance on poor entropy estimators. Finally, we end our paper with concrete recommendations for entropy estimation depending on distribution type and data availability.
When generating natural language from neural probabilistic models, high probability does not always coincide with high quality: It has often been observed that mode-seeking decoding methods, i.e., those that produce high-probability text under the model, lead to unnatural language. On the other hand, the lower-probability text generated by stochastic methods is perceived as more human-like. In this note, we offer an explanation for this phenomenon by analyzing language generation through an information-theoretic lens. Specifically, we posit that human-like language should contain an amount of information (quantified as negative log-probability) that is close to the entropy of the distribution over natural strings. Further, we posit that language with substantially more (or less) information is undesirable. We provide preliminary empirical evidence in favor of this hypothesis; quality ratings of both human and machine-generated text -- covering multiple tasks and common decoding strategies -- suggest high-quality text has an information content significantly closer to the entropy than we would expect by chance.
Numerous analyses of reading time (RT) data have been implemented -- all in an effort to better understand the cognitive processes driving reading comprehension. However, data measured on words at the end of a sentence -- or even at the end of a clause -- is often omitted due to the confounding factors introduced by so-called "wrap-up effects," which manifests as a skewed distribution of RTs for these words. Consequently, the understanding of the cognitive processes that might be involved in these wrap-up effects is limited. In this work, we attempt to learn more about these processes by examining the relationship between wrap-up effects and information-theoretic quantities, such as word and context surprisals. We find that the distribution of information in prior contexts is often predictive of sentence- and clause-final RTs (while not of sentence-medial RTs). This lends support to several prior hypotheses about the processes involved in wrap-up effects.
When generating text from probabilistic models, the chosen decoding strategy has a profound effect on the resulting text. Yet the properties elicited by various decoding strategies do not always transfer across natural language generation tasks. For example, while mode-seeking methods like beam search perform remarkably well for machine translation, they have been observed to lead to incoherent and repetitive text in story generation. Despite such observations, the effectiveness of decoding strategies is often assessed with respect to only a single task. This work -- in contrast -- provides a comprehensive analysis of the interaction between language generation tasks and decoding strategies. Specifically, we measure changes in attributes of generated text as a function of both decoding strategy and task using human and automatic evaluation. Our results reveal both previously-observed and surprising findings. For example, the nature of the diversity-quality trade-off in language generation is very task-specific; the length bias often attributed to beam search is not constant across tasks.
Despite achieving incredibly low perplexities on myriad natural language corpora, today's language models still often underperform when used to generate text. This dichotomy has puzzled the language generation community for the last few years. In this work, we posit that the abstraction of natural language as a communication channel (\`a la Shannon, 1948) can provide new insights into the behaviors of probabilistic language generators, e.g., why high-probability texts can be dull or repetitive. Humans use language as a means of communicating information, and do so in an efficient yet error-minimizing manner, choosing each word in a string with this (perhaps subconscious) goal in mind. We propose that generation from probabilistic models should mimic this behavior. Rather than always choosing words from the high-probability region of the distribution--which have a low Shannon information content--we sample from the set of words with an information content close to its expected value, i.e., close to the conditional entropy of our model. This decision criterion can be realized through a simple and efficient implementation, which we call typical sampling. Automatic and human evaluations show that, in comparison to nucleus and top-k sampling, typical sampling offers competitive performance in terms of quality while consistently reducing the number of degenerate repetitions.
While there exist scores of natural languages, each with its unique features and idiosyncrasies, they all share a unifying theme: enabling human communication. We may thus reasonably predict that human cognition shapes how these languages evolve and are used. Assuming that the capacity to process information is roughly constant across human populations, we expect a surprisal--duration trade-off to arise both across and within languages. We analyse this trade-off using a corpus of 600 languages and, after controlling for several potential confounds, we find strong supporting evidence in both settings. Specifically, we find that, on average, phones are produced faster in languages where they are less surprising, and vice versa. Further, we confirm that more surprising phones are longer, on average, in 319 languages out of the 600. We thus conclude that there is strong evidence of a surprisal--duration trade-off in operation, both across and within the world's languages.