activity
20242026
collaborators

11 papers

cs.AI2026

Beyond the Hivemind: Escaping LLM Homogeneity via Meta-Persona Anchoring and Sequential Temperature Scaling

Tairan Fu, Javier Conde, Carlos Arriaga +5

Recent studies have identified an ``Artificial Hivemind'' effect in Large Language Models (LLMs) causing models to converge on a narrow, homogenized consensus even for open questio…

cs.CL2026

Lost in Sampling: Assessing Lexical Reachability in LLMs via the Word Coverage Score (WCS)

Samer Awad, Javier Conde, Carlos Arriaga +4

Modern Large Language Models (LLMs) are often criticized for producing repetitive and homogeneous text, despite possessing vast latent vocabularies. While previous research has foc…

cs.CL2026

Multiple Choice Questions: Reasoning Makes Large Language Models (LLMs) More Self-Confident, Especially When They are Wrong

Tairan Fu, Javier Conde, Gonzalo Martínez +2

Multiple Choice Question (MCQ) tests are among the most used methods for evaluating large language models (LLMs). Besides checking the correctness of the selected answer, evaluatio…

cs.CV2026

Lost in the Vibrations: Vision Language Models Fail the Dynamic Gauges Test

Tairan Fu, Francisco Javier Santos-Martín, Javier Conde +2

The digital transformation of industrial manufacturing increasingly relies on the ability of autonomous robots to interact with legacy infrastructure, particularly analog gauges. W…

cs.CL2026

Large Language Models and Book Summarization: Reading or Remembering, Which Is Better?

Tairan Fu, Javier Conde, Pedro Reviriego +3

Summarization is a core task in Natural Language Processing (NLP). Recent advances in Large Language Models (LLMs) and the introduction of large context windows reaching millions o…

cs.AI2026

Beyond Reproducibility: Token Probabilities Expose Large Language Model Nondeterminism

Tairan Fu, Gonzalo Martínez, Javier Conde +4

The execution of Large Language Models (LLMs) has been shown to produce nondeterministic results when run on Graphics Processing Units (GPUs), even when they are configured to prod…