The paper investigates whether modestly sized open-source language models can grasp the semantics of rare Paired-Focus constructions, finding that understanding emerges later in training and correlate…

This paper systematically evaluates LLMs' ability to infer pragmatic meaning from non-verbal responses, finding that their accuracy significantly drops compared to verbal inputs.

The paper introduces a multilingual benchmark (MentalMap) to test if LLMs build internal spatial world models from text, finding a universal 'L3 reasoning cliff' suggesting that text-only working memo…

This paper localizes the attention heads within LLMs responsible for specific reasoning steps, finding that specialized heads handle factual retrieval while higher layers manage global information int…

Weak self-training on synthetic data can amplify a language model's existing capabilities, but this effect is strictly dependent on the compatibility between the source and student models, not on the…

This paper systematically investigates the difficulty of Chinese Zero Pronouns (ZPs) for various LLMs, concluding that ZPs remain a significant and persistent challenge, with state-of-the-art models p…

This paper proposes a domain-specialized large language model, PoetryQwen, for precise translation and emotional understanding of classical poetry.

The study finds that for a relational intervention to successfully restore a language model's behavior after functional collapse, both a relational structure (e.g., acknowledgment) and a first-person…

The paper introduces RefMem-Bench, a new benchmark for measuring reflective memory in long-horizon dialogue, and proposes REMIND, a framework that significantly improves models' ability to synthesize…

The paper introduces a Behavioral Specification, an interpretive layer that significantly improves AI personalization by measuring and maximizing 'representational accuracy'—how well the AI captures t…

The paper identifies five persistent, deep-seated behavioral patterns ('training strata') in LLMs, observed through long-term, intimate human-AI interaction, suggesting that training artifacts survive…

The paper introduces a novel framework to quantify faithful confidence expression (FC) in Large Reasoning Models (LRMs), finding that FC remains a significant and challenging reliability target for th…

This paper evaluates the performance of a Large Language Model (LLM) in a high-stakes context by comparing it to human experts and measuring variance and error magnitude.

This paper evaluates the performance of a Large Language Model (LLM) in a high-stakes context by comparing it to human experts and measuring variance and error magnitude.

The paper introduces a new quantitative metric, Contextual Alternative Choice (CAC), to rigorously test language models' syntactic and functional understanding of determiners, showing that current mod…

The study demonstrates that domain adaptation primarily reshapes the linguistic explanatory framework of language models, causing shifts in cosmological stance secondarily, rather than directly modify…

The paper demonstrates that the AI-like style introduced by post-training alignment can be measured, localized, and causally removed using a novel ablation technique called PASTA.

The paper proposes a Multi-Phase Inference Mechanism (MIM) to formalize how diverse world models arise, reframing alignment as making heterogeneous representations mutually processable rather than for…

The paper introduces MIDI, a novel multilingual dataset that embeds idioms in realistic sentence and conversational contexts across diverse resource levels, revealing that idiom comprehension is signi…