Papers by Reto Gubelmann
Pragmatic Norms Are All You Need – Why The Symbol Grounding Problem Does Not Apply to LLMs (2024.emnlp-main)
Copied to clipboard
| Challenge: | 'Symbol grounding problem' is a philosophical problem that arises when questionable theories of meaning are presupposed. |
| Approach: | They argue that LLMs are vulnerable to Harnad’s symbol grounding problem (SGP), as it has been claimed recently . they trace the origins of the SGP to the computational theory of mind . |
| Outcome: | The proposed model-theoretic semantics does not give rise to the SGP, as it has been claimed in the literature. |
Assessing Reliability and Political Bias In LLMs’ Judgements of Formal and Material Inferences With Partisan Conclusions (2025.acl-long)
Copied to clipboard
| Challenge: | This paper examines the ability of LLMs to correctly label simple inferences with partisan conclusions. |
| Approach: | They develop a dataset with formal and material inferences with conclusions that favor either the political left or the political right. |
| Outcome: | The proposed models show that they are unreliable and political bias persists throughout the English and German datasets. |
Too Fast, Too Shallow – LLMs, Including Reasoning LLMs, Are Unreliable Constitutional Reasoners (2026.findings-acl)
Copied to clipboard
| Challenge: | Using three different datasets, we assess LLMs’ constitutional reasoning abilities using three different constitutional frameworks. |
| Approach: | They propose to use the influential dual process theory of cognition to assess LLMs' constitutional reasoning abilities. |
| Outcome: | The LLMs label less than 70% correctly and open-weight reasoning LLM and gpt-4o outperform open- weight non-reasoning LLM. |
When Truth Matters - Addressing Pragmatic Categories in Natural Language Inference (NLI) by Large Language Models (LLMs) (2023.starsem-1)
Copied to clipboard
| Challenge: | In this paper, we examine the ability of large language models (LLMs) to accommodate different pragmatic sentence types, such as questions, commands, and sentence fragments for natural language inference (NLI). |
| Approach: | They propose to fine-tune large language models to accommodate different sentence types for natural language inference (NLI) they also explore ChatGPT's concept of entailment by using a symbolic semantic parser. |
| Outcome: | The proposed models can accommodate different sentence types without losing too much accuracy on MNLI-matched models. |
The Shift from Logic to Dialectic in Argumentation Theory: Implications for Computational Argument Quality Assessment (2025.coling-main)
Copied to clipboard
| Challenge: | In the field of computational argument quality assessment, logic and dialectic are essential dimensions used to measure the quality of argumentative texts. |
| Approach: | They propose to separate logic and dialectic as quality dimensions in computational argument quality assessment . they propose to use dialectical considerations to improve the quality of argumentative texts . |
| Outcome: | The proposed method can benefit argument theory and argument analysis by separating the two quality dimensions. |
Context Matters: A Pragmatic Study of PLMs’ Negation Understanding (2022.acl-long)
Copied to clipboard
| Challenge: | In linguistics, there are two main perspectives on negation: a semantic and a pragmatic view. |
| Approach: | They propose to use transformer-based pre-trained language models to study negation understanding using a pragmatic paradigm. |
| Outcome: | The proposed transformer-based model outperforms the human benchmark at NLU and GLUE, and the results are much more optimistic than previous studies. |
Sentence Smith: Controllable Edits for Evaluating Text Embeddings (2025.emnlp-main)
Copied to clipboard
| Challenge: | Controllable and transparent text generation has been a long-standing goal in NLP . but previous approaches were hindered by parsing and generation insufficiencies . |
| Approach: | They propose a framework for English that has three steps: 1. Parsing a sentence into a semantic graph. 2. Applying human-designed semantic manipulation rules. 3. Generating text from the manipulated graph. |
| Outcome: | The proposed framework for English is based on a neural network and parsers. |