We present a mechanistic interpretability study of GPT-2 that causally examines how sentiment information is processed across its transformer layers. Using systematic activation patching across all 12 layers, we test the hypothesized two-stage sentiment architecture comprising early lexical detection and mid-layer contextual integration. Our experiments confirm that early layers (0-3) act as lexical sentiment detectors, encoding stable, position specific polarity signals that are largely independent of context. However, all three contextual integration hypotheses: Middle Layer Concentration, Phenomenon Specificity, and Distributed Processing are falsified. Instead of mid-layer specialization, we find that contextual phenomena such as negation, sarcasm, domain shifts etc. are integrated primarily in late layers (8-11) through a unified, non-modular mechanism. These experimental findings provide causal evidence that GPT-2's sentiment computation differs from the predicted hierarchical pattern, highlighting the need for further empirical characterization of contextual integration in large language models.

Human-like personality traits have recently been discovered in large language models, raising the hypothesis that their (known and as yet undiscovered) biases conform with human latent psychological constructs. While large conversational models may be tricked into answering psychometric questionnaires, the latent psychological constructs of thousands of simpler transformers, trained for other tasks, cannot be assessed because appropriate psychometric methods are currently lacking. Here, we show how standard psychological questionnaires can be reformulated into natural language inference prompts, and we provide a code library to support the psychometric assessment of arbitrary models. We demonstrate, using a sample of 88 publicly available models, the existence of human-like mental health-related constructs (including anxiety, depression, and Sense of Coherence) which conform with standard theories in human psychology and show similar correlations and mitigation strategies. The ability to interpret and rectify the performance of language models by using psychological tools can boost the development of more explainable, controllable, and trustworthy models.

Several methodologies have recently been proposed to evaluate the ability of Pretrained Language Models (PLMs) to interpret negation. In this article, we build on Gubelmann and Handschuh (2022), which studies the modification of PLMs' predictions as a function of the polarity of inputs, in English. Crucially, this test uses ``self-contained'' inputs ending with a masked position: depending on the polarity of a verb in the input, a particular token is either semantically ruled out or allowed at the masked position. By replicating Gubelmann and Handschuh (2022) experiments, we have uncovered flaws that weaken the conclusions that can be drawn from this test. We thus propose an improved version, the Self-Contained Neg Test, which is more controlled, more systematic, and entirely based on examples forming minimal pairs varying only in the presence or absence of verbal negation in English. When applying our test to the roberta and bert base and large models, we show that only roberta-large shows trends that match the expectations, while bert-base is mostly insensitive to negation. For all the tested models though, in a significant number of test instances the top-1 prediction remains the token that is semantically forbidden by the context, which shows how much room for improvement remains for a proper treatment of the negation phenomenon.

Certain forms of linguistic annotation, like part of speech and semantic tagging, can be automated with high accuracy. However, manual annotation is still necessary for complex pragmatic and discursive features that lack a direct mapping to lexical forms. This manual process is time-consuming and error-prone, limiting the scalability of function-to-form approaches in corpus linguistics. To address this, our study explores automating pragma-discursive corpus annotation using large language models (LLMs). We compare ChatGPT, the Bing chatbot, and a human coder in annotating apology components in English based on the local grammar framework. We find that the Bing chatbot outperformed ChatGPT, with accuracy approaching that of a human coder. These results suggest that AI can be successfully deployed to aid pragma-discursive corpus annotation, making the process more efficient and scalable. Keywords: linguistic annotation, function-to-form approaches, large language models, local grammar analysis, Bing chatbot, ChatGPT

Sarcasm is an advanced linguistic expression often found on various online platforms. Sarcasm detection is challenging in natural language processing tasks that affect sentiment analysis. This article presents the inventive method of the semigraph, including semigraph construction and sarcasm detection processes. A variation of the semigraph is suggested in the pattern-relatedness of the text document. The proposed method is to obtain the sarcastic and non-sarcastic polarity scores of a document using a semigraph. The sarcastic polarity score represents the possibility that a document will become sarcastic. Sarcasm is detected based on the polarity scoring model. The performance of the proposed model enhances the existing prior art approach to sarcasm detection. In the Amazon product review, the model achieved the accuracy, recall, and f-measure of 0.87, 0.79, and 0.83, respectively.