Expect the Unexpected? Testing the Surprisal of Salient Entities

• URL: http://arxiv.org/abs/2604.10724v1
• Date: Sun, 12 Apr 2026 16:52:05 GMT
• Title: Expect the Unexpected? Testing the Surprisal of Salient Entities
• Abstract summary: We study how overall salience of entities in discourse relates to surprisal using 70K manually annotated mentions.<n>Our results show that globally salient entities exhibit significantly higher surprisal than non-salient ones, even controlling for position, length, and nesting confounds.<n>Our findings refine the UID competing pressures framework by identifying global entity salience as a mechanism shaping information distribution in discourse.
• Score: 8.185354486103291
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Previous work examining the Uniform Information Density (UID) hypothesis has shown that while information as measured by surprisal metrics is distributed more or less evenly across documents overall, local discrepancies can arise due to functional pressures corresponding to syntactic and discourse structural constraints. However, work thus far has largely disregarded the relative salience of discourse participants. We fill this gap by studying how overall salience of entities in discourse relates to surprisal using 70K manually annotated mentions across 16 genres of English and a novel minimal-pair prompting method. Our results show that globally salient entities exhibit significantly higher surprisal than non-salient ones, even controlling for position, length, and nesting confounds. Moreover, salient entities systematically reduce surprisal for surrounding content when used as prompts, enhancing document-level predictability. This effect varies by genre, appearing strongest in topic-coherent texts and weakest in conversational contexts. Our findings refine the UID competing pressures framework by identifying global entity salience as a mechanism shaping information distribution in discourse.

Related papers

• Inference-Time Steering for Cross-Lingual Factual Consistency in LLMs [51.56484100374058]
  Large Language Models (LLMs) demonstrate remarkable multilingual fluency, but their internal knowledge representations remain disproportionately biased toward high-resource languages.<n>We investigate whether these biases can be mitigated at inference time, forcing an English-prompted model to answer as if it were queried in target languages.<n>Experiments on Gemma 3 12B Instruct reveal persona prompting to be the strongest overall intervention, balancing efficacy, safety, and out-of-domain generalization.
  arXiv  Detail & Related papers  (2026-07-21T16:15:05Z)
• Textual Belief States for World Models: Identifiable Representation Learning Under Strict Mediation [19.960113437887383]
  We show how to make strict mediation work in the text domain.<n>We then introduce textual latent states, which are discrete, interpretable, and variable-length.<n>Experiments on TextWorld and ScienceWorld show preserved one-step prediction accuracy alongside up to 57% gains in representation quality.
  arXiv  Detail & Related papers  (2026-06-26T03:26:02Z)
• Farther the Shift, Sparser the Representation: Analyzing OOD Mechanisms in LLMs [100.02824137397464]
  We investigate how Large Language Models adapt their internal representations when encountering inputs of increasing difficulty.<n>We reveal a consistent and quantifiable phenomenon: as task difficulty increases, the last hidden states of LLMs become substantially sparser.<n>This sparsity--difficulty relation is observable across diverse models and domains.
  arXiv  Detail & Related papers  (2026-03-03T18:48:15Z)
• Is Information Density Uniform when Utterances are Grounded on Perception and Discourse? [7.3258783042969675]
  We present the first computational study of Uniform Information Density (UID) in visually grounded settings.<n>We find that grounding on perception consistently smooths the distribution of information, increasing both global and local uniformity.<n>Overall, this study takes a first step towards modelling the temporal dynamics of information flow in ecologically plausible, multimodal language use.
  arXiv  Detail & Related papers  (2026-02-16T11:25:00Z)
• The Harmonic Structure of Information Contours [54.38365999922221]
  We find consistent evidence of periodic patterns in information rate in texts in English, Spanish, German, Dutch, Basque, and Brazilian Portuguese.<n>Many dominant frequencies align with discourse structure, suggesting these oscillations reflect meaningful linguistic organization.
  arXiv  Detail & Related papers  (2025-06-04T12:56:30Z)
• Surprise! Uniform Information Density Isn't the Whole Story: Predicting Surprisal Contours in Long-form Discourse [54.08750245737734]
  We propose that speakers modulate information rate based on location within a hierarchically-structured model of discourse. We find that hierarchical predictors are significant predictors of a discourse's information contour and that deeply nested hierarchical predictors are more predictive than shallow ones.
  arXiv  Detail & Related papers  (2024-10-21T14:42:37Z)
• On the Role of Context in Reading Time Prediction [50.87306355705826]
  We present a new perspective on how readers integrate context during real-time language comprehension.<n>Our proposals build on surprisal theory, which posits that the processing effort of a linguistic unit is an affine function of its in-context information content.
  arXiv  Detail & Related papers  (2024-09-12T15:52:22Z)
• JoPA:Explaining Large Language Model's Generation via Joint Prompt Attribution [41.89180735617241]
  Large Language Models (LLMs) have demonstrated impressive performances in complex text generation tasks.<n>The contribution of the input prompt to the generated content still remains obscure to humans.<n>We introduce a counterfactual explanation framework based on Joint Prompt Attribution, JoPA.
  arXiv  Detail & Related papers  (2024-05-30T18:16:41Z)
• Revisiting the Uniform Information Density Hypothesis [44.277066511088634]
  We investigate the uniform information density (UID) hypothesis using reading time and acceptability data. For acceptability judgments, we find clearer evidence that non-uniformity in information density is predictive of lower acceptability.
  arXiv  Detail & Related papers  (2021-09-23T20:41:47Z)
• Did the Cat Drink the Coffee? Challenging Transformers with Generalized Event Knowledge [59.22170796793179]
  Transformers Language Models (TLMs) were tested on a benchmark for the textitdynamic estimation of thematic fit Our results show that TLMs can reach performances that are comparable to those achieved by SDM. However, additional analysis consistently suggests that TLMs do not capture important aspects of event knowledge.
  arXiv  Detail & Related papers  (2021-07-22T20:52:26Z)
• Pareto Probing: Trading Off Accuracy for Complexity [87.09294772742737]
  We argue for a probe metric that reflects the fundamental trade-off between probe complexity and performance. Our experiments with dependency parsing reveal a wide gap in syntactic knowledge between contextual and non-contextual representations.
  arXiv  Detail & Related papers  (2020-10-05T17:27:31Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.