United States

This paper investigates good-enough parsing in Turkish by comparing human self-paced reading performance to the surprisal and attention patterns of three Turkish Large Language Models (LLMs), GPT-2-Base, GPT-2-Large, and LLaMA-3. The results show that Turkish speakers rely on good-enough parsing for implausible but grammatically permissible sentences (e.g., interpreting sentences such as `the man bit the dog&#39; as &#39;the dog bit the man&#39;). Although the smaller LLMs (e.g., GPT-2) were better predictors of human RTs, they seem to have relied more heavily on semantic plausibility than humans. Comparably, larger LLMs (e.g., LLaMA-3) tended to make more probabilistic parsing based on word order, exhibiting less *good-enough* parsing behavior. Therefore, we conclude that LLMs take syntactic and semantic constraints into account when processing thematic roles, but not to the same extent as human parsers.

NAACL 2025

When Men Bite Dogs: Testing Good-Enough Parsing in Turkish with Humans and Large Language Models

good-enough parsing

semantic plausibility

turkish

large language models

This paper investigates good-enough parsing in Turkish by comparing human self-paced reading performance to the surprisal and attention patterns of three Turkish Large Language Models (LLMs), GPT-2-Base, GPT-2-Large, and LLaMA-3. The results show that Turkish speakers rely on good-enough parsing for implausible but grammatically permissible sentences (e.g., interpreting sentences such as `the man bit the dog' as 'the dog bit the man'). Although the smaller LLMs (e.g., GPT-2) were better predictors of human RTs, they seem to have relied more heavily on semantic plausibility than humans. Comparably, larger LLMs (e.g., LLaMA-3) tended to make more probabilistic parsing based on word order, exhibiting less *good-enough* parsing behavior. Therefore, we conclude that LLMs take syntactic and semantic constraints into account when processing thematic roles, but not to the same extent as human parsers.

workshop paper

### Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Welcome to the 2025 meeting of the Nations of the Americas Chapter of the Association for Computational Linguistics! I am proud to help organize the first NAACL conference to carry the new name of our organization, one that emphasizes inclusion for all of the Americas. I am also pleased to welcome you to Albuquerque, New Mexico, a state whose unique blend of cultural influences will make for an excellent backdrop for NAACL 2025, especially with this year’s special theme on NLP in a Multicultural World. 
**[Continue reading...](https://drive.google.com/file/d/1jX-qGhqVSZZCIrAnJaz798pflu5Irrdn/view?usp=sharing)**

*- Colin Cherry, Google, NAACL 2025 General Chair* 

[![](https://assets.underline.io/markdown_image/1/image/b087f8a4dc5816d6e1a6514e59c59ac3.png)](https://drive.google.com/file/d/1T96GzPqObXrMln2BMByCSXTSizjTg69P/view?usp=sharing)

You need to log in with the email address you registered with. Access credentials have been sent to your email. 

Please be sure to check your spam and other email folders if you do not see an email confirmation right away.

Please log in to explore this event.

To access NAACL 2025 event page you are required to register. Please follow [**this link**](https://2025.naacl.org/registration/registration/) to register. Access will depend on your registration type.

Please register!

Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Humans tend to predict the next words during sentence comprehension, but under unique circumstances, they demonstrate an ability for longer coherent word sequence prediction. In this paper, we investigate whether Transformers can model such hyperprediction observed in humans during sentence processing, specifically in the context of Japanese buzzer quizzes. We conducted eye-tracking experiments where the participants read the first half of buzzer quiz questions and predicted the second half, while we modeled their reading time using the GPT-2. By modeling the reading times of each word in the first half of the question using GPT-2 surprisal, we examined under what conditions fine-tuned language models can better predict reading times. As a result, we found that GPT-2 surprisal effectively explains the reading times of quiz experts as they read the first half of the question while predicting the latter half. When the language model was fine-tuned with quiz questions, the perplexity value decreased. Lower perplexity corresponded to higher psychometric predictive power; however, excessive data for fine-tuning led to a decrease in perplexity and the fine-tuned model exhibited a low psychometric predictive power. Overall, our findings suggest that a moderate amount of data is required for fine-tuning in order to model human hyperprediction.

Transformers Can Model Human Hyperprediction in Buzzer Quiz

This study examines how sentence structure shapes contrast predictions in both humans and large language models (LLMs). Using Mandarin ditransitive constructions — double object (DO, “She gave the girl the candy, but not…”) vs. prepositional object (PO, “She gave the candy to the girl, but not…”) as a testbed, we employed a sentence continuation task involving three human groups (written, spoken, and prosodically normalized spoken stimuli) and three LLMs (GPT-4o, LLaMA-3, and Qwen-2.5). Two principal findings emerged: (1) Although human participants predominantly focused on the theme (e.g., “the candy”), contrast predictions were significantly modulated by sentence structure—particularly in spoken contexts, where the sentence-final element drew more attention. (2) While LLMs showed a similar reliance on structure, they displayed a larger effect size and more closely resembled human spoken data than written data, indicating a stronger emphasis on linear order in generating contrast predictions. By adopting a unified psycholinguistic paradigm, this study advances our understanding of predictive language processing for both humans and LLMs and informs research on human–model alignment in linguistic tasks.

What to Predict? Exploring How Sentence Structure Influences Contrast Predictions in Humans and Large Language Models

This paper uses autoregressive large language models to explore at which points in a given input sentence the semantic information is decodable. Using representational similarity analysis and probing, the results show that autoregressive models are capable of extracting the semantic relation information from a dataset of noun-noun compounds. When considering the effect of processing the head and modifier nouns in context, the extracted representations show greater correlation after processing both constituent nouns in the same sentence. The linguistic properties of the head nouns may influence the ability of LLMs to extract relation information when the head and modifier words are processed separately. Probing suggests that Phi-1 and LLaMA-3.2 are exposed to relation information during training, as they are able to predict the relation vectors for compounds from separate word representations to a similar degree as using compositional compound representations. However, the difference in processing condition for GPT-2 and DeepSeek-R1 indicates that these models are actively processing the contextual semantic relation information of the compound.

Investigating noun-noun compound relation representations in autoregressive large language models

Large language models (LLMs) exhibit remarkable capabilities on not just language tasks, but also various tasks that are not linguistic in nature, such as logical reasoning and social inference. In the human brain, neuroscience has identified a core language system that selectively and causally supports language processing. We here ask whether similar specialization for language emerges in LLMs. We identify language-selective units within 18 popular LLMs, using the same localization approach that is used in neuroscience. We then establish the causal role of these units by demonstrating that ablating LLM language-selective units -- but not random units -- leads to drastic deficits in language tasks. Correspondingly, language-selective LLM units are more aligned to brain recordings from the human language system than random units. Finally, we investigate whether our localization method extends to other cognitive domains: while we find specialized networks in some LLMs for reasoning and social capabilities, there are substantial differences among models. These findings provide functional and causal evidence for specialization in large language models, and highlight parallels with the functional organization in the brain.

The LLM Language Network: A Neuroscientific Approach for Identifying Causally Task-Relevant Units

Pragmatic accommodation is a key factor in maintaining smooth real-life communication, yet it is largely overlooked in pragmatic reasoning models such as the Rational Speech Act (RSA) model. The current study explores ways of extending the basic RSA model: revising the listener’s belief of common ground, adjusting the listener’s belief of speaker rationality, and doing both simultaneously. We evaluated model predictions for utterances varying in utility levels (i.e., how useful an utterance is for state updating) and sentence polarities. We find that (i) contra prior findings, low utility does not always trigger the expected extra inferences, but high utility does, (ii) higher utility is associated with higher speaker rationality, and (iii) our combined model predicts that lower speaker rationality and lower wonkiness co-occur. Theoretical implications of these findings are discussed.

Modeling Pragmatic Accommodation: A [Rational, Irrational] Speaker in a [Wonky, Normal] World

The achievements of Large Language Models in Natural Language Processing, especially for high-resource languages, call for a better understanding of their characteristics from a cognitive perspective.
Researchers have attempted to evaluate artificial models by testing their ability to predict behavioral (e.g., eye-tracking fixations) and physiological (e.g., brain responses) variables during language processing (e.g., reading/listening). In this paper, we propose using spontaneous speech corpora to derive production variables (speech reductions, prosodic prominences) and applying them in a similar fashion. More precisely, we extract. We then test models trained with a standard procedure on different pretraining datasets (written, spoken, and mixed genres) for their ability to predict these two variables. Our results show that, after some fine-tuning, the models can predict these production variables well above baselines. We also observe that spoken genre training data provides more accurate predictions than written genres. These results contribute to the broader effort of using high-quality speech corpora as benchmarks for LLMs.

Spontaneous Speech Variables as a Test of Human Cognitive Adequacy in LLMs

Temporal connectives like before and after play an important role in sentence comprehension by linking events in the sentence to real-world event order. Previous psycholinguistic research has shown that sentences where the order of mention aligns with real-world event order (chronological order) are easier to process than those that do not (reverse order). In this study, we investigate whether BERT, a large language model (LLM), reflects this humanlike preference for chronological order when predicting temporal connectives. We conduct experiments using both ambiguous and unambiguous sentences, evaluating the model’s ability to predict the appropriate connective in sentence-initial and sentence-medial positions. Our results confirm prior findings that BERT prefers chronological order only in sentence-initial positions when sentences are ambiguous. However, we extend this research by demonstrating that BERT also shows a chronological preference in sentence-medial positions when sentences are unambiguous. Additionally, we show that the model achieves higher accuracy when the correct connective reflects chronological order, illustrating its sensitivity to event order. These findings indicate that BERT predictions on event order align with human psycholinguistic patterns, with implications for cognitive modelling and event knowledge representations in LLMs.

Event Order Recognition in BERT: Evidence from Temporal Connectives

With the increasing interest in using large language models (LLMs) for planning in natural language, understanding their behaviors becomes an important research question. This work conducts a systematic investigation of LLMs' ability to perform logical reasoning in natural language. We introduce a controlled dataset of hypothetical and disjunctive syllogisms in propositional and modal logic and use it as the testbed for understanding LLM performance. Our results lead to novel insights in predicting LLM behaviors: in addition to the probability of input, logical forms should be considered as important factors. In addition, we show similarities and discrepancies between the logical reasoning performances of humans and LLMs by collecting and comparing behavioral data from both.

Logical forms complement probability in understanding language model (and human) performance

What is a neural model with minimum architectural complexity that exhibits reasonable language learning capability? To explore such a simple but sufficient neural language model, we revisit a basic reservoir computing (RC) model, Echo State Network (ESN), a restricted class of simple Recurrent Neural Networks. Our experiments showed that ESN with a large hidden state is comparable or superior to Transformer in grammaticality judgment tasks when trained with about 100M words, suggesting that architectures as complex as that of Transformer may not always be necessary for syntactic learning.

Syntactic Learnability of Echo State Neural Language Models at Scale

Recent research has increasingly focused on the extent to which large language models (LLMs) exhibit human-like behavior. In this study, we investigate whether the mental lexicon in LLMs resembles that of humans in terms of lexical organization. Using a word association task—a direct and widely used method for probing word meaning and relationships in the human mind—we evaluated the lexical representations of GPT-4 and Llama-3.1. Our findings reveal that LLMs closely emulate human mental lexicons in capturing semantic relatedness but exhibit notable differences in other properties, such as association frequency and dominant lexical patterns (e.g., top associates). Specifically, LLM lexicons demonstrate greater clustering and reduced diversity compared to the human lexicon, with KL divergence analysis confirming significant deviations in word association patterns. Additionally, LLMs fail to capture word association responses patterns in different demographic human groups. Among the models, GPT-4 consistently exhibited a slightly higher degree of human-likeness than Llama-3.1. This study highlights both the potential and limitations of LLMs in replicating human mental lexicons, offering valuable insights for applications in natural language processing and cognitive science research involving LLMs.

Downloads

Next from NAACL 2025

Transformers Can Model Human Hyperprediction in Buzzer Quiz

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from NAACL 2025

Transformers Can Model Human Hyperprediction in Buzzer Quiz

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads