4arXiv cs.CL (Computation and Language)·
AMALIA-9B study finds LLM annotation agreement does not imply construct validity for moral foundation coding
A new arXiv paper tests whether Portugal's publicly funded AMALIA-9B language model can validly annotate moral foundation constructs (specifically authority) in European Portuguese text, not just agree with human coders. Using a 'recovery gap' methodology — comparing holistic prompting against decomposed codebook-clause prompting — the authors find AMALIA recovers only about half its holistic performance under decomposition, suggesting reliance on surface correlates rather than genuine construct understanding. A larger open multilingual model closes the same gap on the same corpus, pointing to model-level rather than corpus-level failure. The paper argues that sovereign/national LLM benchmark batteries should test the evidential route to agreement, not just agreement rates.
Related guides (3)
Related events (8)
Moral Semantics Survive Machine Translation: Cross-Lingual Evidence from Moral Foundations Corpora
This paper investigates whether LLM-based machine translation can preserve moral semantic content well enough to enable cross-lingual moral values classification, using Polish as a test case with ~50k annotated social media posts. A four-method validation pipeline (LaBSE embedding similarity, CKA, LLM-as-judge, and classifier parity) shows mean cosine similarity of 0.86 and AUC gaps of only 0.01–0.02 across Moral Foundations categories. The results suggest machine translation is a practical path to extending moral values NLP research to under-resourced languages, with expected generalization to related Slavic languages.
Cross-lingual evaluation framework reveals LLMs redistribute cultural narrative structure while preserving semantic meaning
A new arXiv preprint introduces a multilingual evaluation framework using 414 proverbs across 15 languages to assess whether LLMs preserve culturally grounded meaning when generating narratives. Using four LLMs to produce 13k narratives, the study finds that cross-lingual prompting preserves proverb-level semantic meaning but systematically redistributes agency, social positioning, and narrative structure. Strong inter-model convergence across architectures suggests multilingual LLMs rely on shared semantic abstractions. The authors argue that semantic similarity metrics alone overestimate cultural preservation in multilingual evaluations.
LLMs fail to consistently simulate demographic perspective-taking in hate speech annotation
A new arXiv paper evaluates whether persona-conditioned LLMs can replicate how different demographic groups perceive hate speech, testing three dimensions: inter-group disagreement, in-group sensitivity, and vicarious prediction. No model consistently captures all three dimensions, and performance is highly model-dependent rather than emerging reliably from identity prompts alone. Vicarious prompting with Llama 3.1 provides the closest approximation to human disagreement patterns across demographic axes. The findings have implications for using LLMs as proxies for diverse human annotators in content moderation tasks.
Calibrated LLM annotation and encoder transfer for measuring human values in social media text
A new arXiv preprint investigates how different LLMs, prompts, and instruction languages operationalize Schwartz's theory of basic human values when annotating non-English social media posts. The authors evaluate annotation quality beyond standard F1 metrics, examining structural alignment, error structure, and confidence-ambiguity relations, finding that iterative prompt calibration reduces misattributions. They also demonstrate that LLM annotations can be transferred to a smaller encoder model via soft-label training, preserving theory-grounded value interpretations and uncertainty information.
Tracing the Emergence of Human-Like Pragmatic Reasoning in LLMs Across Languages
Researchers conducted a population-matching experiment evaluating 25 LLMs on conditional inference tasks across four languages, comparing model behavior to matched human populations. The study finds that LLMs function as accurate semantic operators but systematically fail to capture pragmatic enrichments—context-sensitive inferences beyond literal logical meaning—that humans apply effortlessly. Model performance on pragmatic reasoning is not predicted by open vs. closed weights, training orientation, or architecture type, suggesting pragmatic reasoning remains an emergent and unreliable capability. The findings contribute to ongoing debates about whether LLMs reason like humans or merely approximate surface-level linguistic patterns.
Towards Reliable Multilingual LLMs-as-a-Judge: An Empirical Study
This paper systematically investigates strategies for extending LLM-based automatic evaluation (LLMs-as-a-Judge) to multilingual settings, covering high-, mid-, and low-resource languages (English, Spanish, Basque). The authors compare instruction translation, monolingual vs. multilingual supervision, and model size, finding that fine-tuned smaller models can match proprietary models when in-domain data is available, while zero-shot larger models are preferable out-of-domain. Two meta-evaluation datasets are extended to Spanish and Basque, and all data and code are publicly released.
BabelJudge: Benchmark for measuring LLM-as-a-judge reliability across languages and agent trajectories
BabelJudge is a new open-source benchmark and audit framework that systematically measures four failure modes in LLM-as-a-judge systems: position bias, verbosity bias, order inconsistency, and cross-lingual degradation. The framework uses a 'gold-labelling by degradation' technique to generate labeled evaluation pairs without human annotation. Evaluation of Qwen2.5-7B-Instruct-4bit across English, Hindi, Arabic, and Swahili reveals severe cross-lingual reliability drops, with Swahili order consistency collapsing to near-random (0.480). The framework is extended to agentic evaluation with nine trajectory-level perturbations and three new metrics, released as a Python package supporting 11 judge backends.
LLMs fail to reliably self-report adversarial prefill attacks, study finds
A new arXiv paper evaluates whether LLMs can recognize that their own prior responses were elicited by adversarial prefill attacks, testing ten open-weight models (3B–70B) across four safety benchmarks. Models claim intent on prefilled responses only 27.3% of the time on average, and introspective signal is largely mediated by refusal-related reasoning. Three LoRA fine-tuning methods (SFT, GRPO, DPO) improve the intention-probe gap but counterintuitively raise attack success rates on most models, suggesting partial and fragile mitigation. The findings raise concerns about the reliability of LLM self-reports in safety-critical contexts.