Constrained Semantic Decompression in LLMs through Persian Proverb-Conditioned Story Generation

• URL: http://arxiv.org/abs/2606.12599v1
• Date: Wed, 10 Jun 2026 18:54:07 GMT
• Title: Constrained Semantic Decompression in LLMs through Persian Proverb-Conditioned Story Generation
• Abstract summary: We introduce the Proverb Aligned Narrative dataset (PAND), pairing proverbs with human-written stories and explicit meanings.<n>Our findings reveal a persistent emphdecompression gap: current language models fail to faithfully instantiate the underlying moral and causal structure encoded in proverbs.<n>Our proposed task naturally extends to other forms of compressed cultural knowledge.
• Score: 5.105644262054024
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Transforming a dense, abstract proverb into an engaging and morally faithful narrative requires deep cultural understanding and robust semantic grounding. We frame this problem as a \emph{constrained semantic decompression} task and study proverb-conditioned story generation as a testbed for abstraction-to-realization in large language models (LLMs). Focusing on Persian, we introduce the Proverb Aligned Narrative Dataset (PAND), pairing proverbs with human-written stories and explicit meanings. By a hybrid evaluation framework that combines human-calibrated LLM-as-a-Judge with structural metrics, we analyze model behavior across multiple prompting regimes. Our findings reveal a persistent \emph{decompression gap}: current LLMs often achieve strong surface-level fluency while failing to faithfully instantiate the underlying moral and causal structure encoded in proverbs. We further show that explicit reasoning and iterative refinement can partially mitigate these failures, suggesting that many decompression errors arise from difficulties in translating abstract meaning into narrative form rather than a complete lack of relevant knowledge. Our proposed task naturally extends to other forms of compressed cultural knowledge.

Related papers

• Do Large Language Models Play Six Degrees of Separation? Measuring Topological Compression in Long-Context Manifolds [1.8190501646350212]
  This work mathematically formalizes how transformers execute abstract reasoning and provides a novel, strictly geometric signature for evaluating factual reliability.<n>We show that factually grounded generations maintain structural integrity with their source context, whereas hallucinations induce severe topological collapse.
  arXiv  Detail & Related papers  (2026-08-18T16:05:25Z)
• The Pragmatic Persona: Discovering LLM Persona through Bridging Inference [6.671933768892354]
  Large Language Models (LLMs) reveal inherent and distinctive personas through dialogue.<n>We propose a novel framework that interprets dialogue through bridging inference.<n>Our approach captures latent semantic links that govern how LLMs organize meaning across turns.
  arXiv  Detail & Related papers  (2026-04-27T06:01:47Z)
• Robust Grounding with MLLMs Against Occlusion and Small Objects via Language-Guided Semantic Cues [58.21371989991432]
  Crowded scenes entail visual challenges, which impair object semantics and degrade grounding performance.<n>We propose a novel method that overcomes such constraints by leveraging Language-Guided Semantic Cues (LGSCs)
  arXiv  Detail & Related papers  (2026-04-27T04:42:03Z)
• Diagnosing CFG Interpretation in LLMs [17.612748633265202]
  Given a novel context-free grammar, can LLMs generate syntactically valid, behaviorally functional, and semantically faithful outputs?<n>We introduce RoboGrid, a framework that disentangles syntax, behavior, and semantics through controlled stress-tests of recursion depth, expression complexity, and surface styles.
  arXiv  Detail & Related papers  (2026-04-22T17:43:05Z)
• Codified Foreshadowing-Payoff Text Generation [67.01182739162142]
  Foreshadowing and payoff are ubiquitous narrative devices through which authors introduce commitments early in a story and resolve them through concrete, observable outcomes.<n>Existing evaluations largely overlook this structural failure, focusing on surface-level coherence rather than the logical fulfillment of narrative setups.<n>We introduce Codified Foreshadowing-Payoff Generation, a novel framework that reframes narrative quality through the lens of payoff realization.
  arXiv  Detail & Related papers  (2026-01-11T19:05:37Z)
• MORABLES: A Benchmark for Assessing Abstract Moral Reasoning in LLMs with Fables [50.29407048003165]
  We present MORABLES, a human-verified benchmark built from fables and short stories drawn from historical literature.<n>The main task is structured as multiple-choice questions targeting moral inference, with carefully crafted distractors that challenge models to go beyond shallow, extractive question answering.<n>Our findings show that, while larger models outperform smaller ones, they remain susceptible to adversarial manipulation and often rely on superficial patterns rather than true moral reasoning.
  arXiv  Detail & Related papers  (2025-09-15T19:06:10Z)
• Drivel-ology: Challenging LLMs with Interpreting Nonsense with Depth [21.092167028989632]
  Drivelology is a linguistic phenomenon characterised by "nonsense with depth"<n>We construct a benchmark dataset of over 1,200+ meticulously curated and diverse examples across English, Mandarin, Spanish, French, Japanese, and Korean.<n>We find that current large language models (LLMs) consistently fail to grasp the layered semantics of Drivelological text.
  arXiv  Detail & Related papers  (2025-09-04T03:58:55Z)
• Fantastic Semantics and Where to Find Them: Investigating Which Layers of Generative LLMs Reflect Lexical Semantics [50.982315553104975]
  We investigate the bottom-up evolution of lexical semantics for a popular large language model, namely Llama2. Our experiments show that the representations in lower layers encode lexical semantics, while the higher layers, with weaker semantic induction, are responsible for prediction. This is in contrast to models with discriminative objectives, such as mask language modeling, where the higher layers obtain better lexical semantics.
  arXiv  Detail & Related papers  (2024-03-03T13:14:47Z)
• Improving Factual Consistency of News Summarization by Contrastive Preference Optimization [65.11227166319546]
  Large language models (LLMs) generate summaries that are factually inconsistent with original articles.<n>These hallucinations are challenging to detect through traditional methods.<n>We propose Contrastive Preference Optimization (CPO) to disentangle the LLMs' propensities to generate faithful and fake content.
  arXiv  Detail & Related papers  (2023-10-30T08:40:16Z)
• Simple Linguistic Inferences of Large Language Models (LLMs): Blind Spots and Blinds [59.71218039095155]
  We evaluate language understanding capacities on simple inference tasks that most humans find trivial. We target (i) grammatically-specified entailments, (ii) premises with evidential adverbs of uncertainty, and (iii) monotonicity entailments. The models exhibit moderate to low performance on these evaluation sets.
  arXiv  Detail & Related papers  (2023-05-24T06:41:09Z)
• Lexically-constrained Text Generation through Commonsense Knowledge Extraction and Injection [62.071938098215085]
  We focus on the Commongen benchmark, wherein the aim is to generate a plausible sentence for a given set of input concepts. We propose strategies for enhancing the semantic correctness of the generated text.
  arXiv  Detail & Related papers  (2020-12-19T23:23:40Z)

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.