← Philipp WickePublications
Full record

Publications

Every publication on record, grouped by year, each with a plain-language note on what it set out to do. The front page carries the same list as a scrollable feed; citation counts are live from Google Scholar.

2026
Decoding Workload and Agreement From EEG During Spoken Dialogue With Conversational AI
L. Mihić Zidar, P. Wicke, P. Bhatia, R. Lutz, M. Klug, T. O. Zander · IEEE Int. Winter Conf. on BCI

Tests whether brainwave readings can reveal a person's mental effort and silent agreement while they chat with an AI, and finds effort easier to detect than agreement.

2025
Red and blue language: Word choices in the Trump and Harris 2024 presidential debate
P. Wicke, M. M. Bolognesi · PLOS ONE

Compares the words Trump and Harris used in their 2024 debate and finds she framed issues as recovery while he stressed crisis and decline.

18cited
Time course mechinterp: Analyzing the evolution of components and knowledge in large language models
A. D. Hakimi, A. Modarressi, P. Wicke, H. Schütze · Findings of ACL

Tracks how an AI language model's inner parts shift from general to specialized jobs while learning facts during training, with facts about places mastered before facts about names.

6cited
SLAyiNG: Towards Queer Language Processing
L. Veloso, L. Hirlimann, P. Wicke, H. Schütze · arXiv

Builds the first labeled collection of queer slang from subtitles, social media, and podcasts so language tools stop mistaking it for hate speech.

1cited
Testing Spatial Intuitions of Humans and Large Language and Multimodal Models in Analogies
I. Bueno, A. Bavaresco, J. M. Cunha, P. Wicke · Analogy-Angle II Workshop at ACL 2025

Asks people and AI models to link words with spatial directions through analogies, and finds the models reason about space quite differently from humans.

Exploring LLM's Interpretation of Gestures Within Textual Contexts for Neuroadaptive Technologies
P. Wicke · Neuroadaptive Technology Conference

Tests whether two Llama language models can work out what 96 described hand gestures mean from text alone, and finds they mostly failed.

2024
Exploring Spatial Schema Intuitions in Large Language and Vision Models
P. Wicke, L. Wachowiak · Findings of ACL

Reruns three psychology experiments on AI language and image models to check whether they share people's instincts about basic spatial ideas like up, down, and inside.

12cited
Probing Language Models' Gesture Understanding for Enhanced Human-AI Interaction
P. Wicke · arXiv

Checks whether language models such as GPT-3 and Llama-2 can read hand gestures described in words the way people from different cultures do.

6cited
Using Analogical Reasoning to Prompt LLMs for their Intuitions of Abstract Spatial Schemas
P. Wicke, L. Hirlimann, J. M. Cunha · IJCAI Workshop on Analogical Abstraction

Tests whether prompting large language models with analogies makes their guesses about which words feel upward or downward match human choices more closely.

1cited
Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics
T. Kuribayashi, G. Rambelli, E. Takmaz, P. Wicke, Y. Oseki · CMCL Workshop 2024

Collects the papers from a 2024 workshop where scientists compared how computers and human minds process language, from reading speed to grammar judgments.

Robustness Testing of Multi-Modal Models in Varied Home Environments for Assistive Robots
L. Hirlimann, S. Zhang, H. Schütze, P. Wicke · Geriatronics Summit

Tests how home-helper robot models cope inside a simulated house when the lights are dimmed or walls turned to mirrors, which trips up their vision and movement.

Proceedings of the 2nd Conference on AI for People: Democratizing AI
M. Ziosi, G. Sartor, J. M. Cunha, A. Trotta, P. Wicke · Conference on AI for People, CAIP

Collects eleven peer-reviewed papers from a 2023 Bologna conference on making artificial intelligence more accountable to ordinary people, covering its ethics, governance, law, and privacy.

2023
LoHoRavens: A Long-Horizon Language-Conditioned Benchmark for Robotic Tabletop Manipulation
S. Zhang, P. Wicke, L. K. Şenel, L. Figueredo, A. Naceri, S. Haddadin, B. Plank, H. Schütze · NeurIPS Robot Learning Workshop

Builds a simulated block-stacking test where a robot arm follows multi-step plain-language orders requiring counting, comparing colors, and remembering earlier moves.

23cited
A crosslingual investigation of conceptualization in 1335 languages
Y. Liu, H. Ye, L. Weissweiler, P. Wicke, R. Pei, R. Zangenfeind, H. Schütze · arXiv

Compares how 1,335 languages carve the world into words and shows concrete things line up across languages better than abstract ones, roughly grouping languages into families.

20cited
Immune moral models? Pro-social rule breaking as a moral enhancement approach for ethical AI
R. Ramanayake, P. Wicke, V. Nallur · AI & Society

Argues that helpful AI should sometimes break its designers' rules for the common good, and tests the idea by asking people to solve a COVID vaccine dilemma.

20cited
LMs stand their ground: Investigating the effect of embodiment in figurative language interpretation by language models
P. Wicke · Findings of ACL

Tests whether language models read metaphors better when the actions involve the body, and finds bigger models handle physical metaphors more accurately.

11cited
Towards language-based modulation of assistive robots through multimodal models
P. Wicke, L. K. Şenel, S. Zhang, L. Figueredo, A. Naceri, S. Haddadin, H. Schütze · Geriatronics Summit

Proposes letting elderly-care robots adjust their actions on the fly from spoken commands, and outlines a crowd-sourced set of household instructions to train them.

2cited
2022
The Role of Gestures and Movement in Computational, Embodied Storytelling.
P. Wicke · TREPHAC Workshop at ICCC 2022

Argues that storytelling robots should add hand gestures and body movement to spoken tales, so audiences can follow the action and grasp its meaning more easily.

2cited
2021
Covid-19 discourse on twitter: How the topics, sentiments, subjectivity, and figurative frames changed over time
P. Wicke, M. M. Bolognesi · Frontiers in Communication

Follows COVID-19 tweets through the first 2020 wave, tracking how topics, mood, and figurative frames shifted as lockdowns wore on, with the war metaphor slowly losing ground.

113cited
Creative action at a distance: A conceptual framework for embodied performance with robotic actors
P. Wicke, T. Veale · Frontiers in Robotics and AI

Proposes a way for robots to act out machine-written stories with gestures, poses, and movement, sometimes performing actions metaphorically rather than literally to seem more creative.

13cited
Are you not entertained? computational storytelling with non-verbal interaction
P. Wicke, T. Veale · ACM/IEEE HRI 2021 Companion

Shows a setup where robots tell an AI-generated story and change the plot on the fly from the audience's gestures and facial reactions instead of spoken words.

8cited
Walk the line: Digital storytelling as embodied spatial performance
P. Wicke, T. Veale · AISB 2021 (CC2021)

Turns a spoken story into physical motion, with a performer walking across a space so their shifting positions and gestures mirror how characters relate and the plot moves.

4cited
Metaphor, Blending and Irony in Action: Creative Performance as Interpretation and Emotionally-Grounded Choice.
T. Veale, P. Wicke · ICCC

Gives a robot actor emotional reactions to its lines, letting it interpret a script and choose its own metaphors, blends, and ironic touches rather than merely reciting.

2cited
Computational storytelling as an embodied robot performance with gesture and spatial metaphor
P. Wicke · PhD thesis, University College Dublin

Builds a system that lets robots act out computer-generated stories, using hand gestures and where they stand to convey characters' feelings and relationships.

1cited
2020
Framing COVID-19: How we conceptualize and discuss the pandemic on Twitter
P. Wicke, M. M. Bolognesi · PLOS ONE

Reads COVID-19 tweets from spring 2020 and finds people often described the virus as a war, though the war framing fit some topics better than others.

515cited
Emoji-based semantic representations for abstract and concrete concepts
P. Wicke, M. Bolognesi · Cognitive Processing

Gathered the emoji people link to 300 English nouns and used them as picture-based meaning cues to compare how abstract and concrete words differ.

40cited
The show must go on: On the use of embodiment, space and gesture in computational storytelling
P. Wicke, T. Veale · New Generation Computing

Uses pairs of robots to act out computer-written stories, with pantomime gestures and stage positioning standing in for characters' emotions and social ties.

16cited
Show, Don't (Just) Tell: Embodiment and Spatial Metaphor in Computational Story-Telling.
P. Wicke, T. Veale · ICCC

Programs two small humanoid robots and an Alexa to act out a story with gestures and stage movement, so they perform it rather than just narrate.

5cited
An Approach for Text-to-Emoji Translation
P. Wicke, J. M. Cunha · ICCC 2020

Builds a system that turns sentences from tweets, stories, and poems into emoji using word-meaning lookups instead of machine learning, then compares it with emoji-prediction apps.

2cited
2019
Duets Ex Machina: On The Performative Aspects of" Double Acts" in Computational Creativity.
T. Veale, P. Wicke, T. Mildner · ICCC

Pairs an Alexa voice storyteller with a NAO robot that acts out the tale, turning the friction between the chatty one and the clumsy one into comedy.

11cited
Storytelling with Alexa and Nao
P. Wicke, T. Veale · Dagstuhl Seminar 19172

Pairs an Amazon Alexa speaker with a NAO robot so the two perform a story together as a comedy double act, each covering the other's weaknesses.

2018
Interview with the Robot: Question-Guided Collaboration in a Storytelling System.
P. Wicke, T. Veale · ICCC

Builds a NAO robot that interviews someone about their life, then turns the answers into new stories it tells back using gestures.

20cited
Storytelling by a Show of Hands: A framework for interactive embodied storytelling in robotic agents
P. Wicke, T. Veale · AISB 2018

Builds a robot that acts out stories by matching each of about 800 plot verbs to a hand gesture, and finds the gestures made audiences laugh more.

10cited
Wheels Within Wheels: A Causal Treatment of Image Schemas in An Embodied Storytelling System.
P. Wicke, T. Veale · TriCoLore (C3GI/ISD/SCORE)

Builds a Nao robot that tells stories aloud while picking hand gestures from nine recurring spatial patterns, linking what happens in the plot to body movement.

6cited
2017
Ideograms as semantic primes: Emoji in computational linguistic creativity
P. Wicke · BSc thesis, Osnabrück University

Builds a system that turns words, especially action verbs, into strings of emoji using picture-puzzle tricks, metaphor, and idioms.

14cited