Author

Roland Memisevic

h-index 5 View on Semantic Scholar

We have 2 of 21 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Preprint Sep 2026

Alleviating Hallucination in Reasoning Tasks with Training-Free Uncertainty-Guided Steering

USteer, a simple, training-free steering mechanism that adjusts a model's layer-wise activations during inference using the gradient of a confidence measure with respect to the activations, demonstrates that confidence signals can be leveraged not only for detection, but also for effective inference-time control of mod...

Li-Tian Liu, Qi-Qi Hou, Yu-Bin Jian et al. · 0 citations

Jul 2026

On Locality and Length Generalization in Visual Reasoning

The experiments reveal that, similar to language models, vision models can learn to exploit global shortcuts and thereby fail to generalize over task length or complexity, but it is shown that recurrent vision policies based on strictly local perception can mitigate these failures, thereby allowing models to generalize...

Pulkit Madan, S. Haresh, Reza Ebrahimi et al. · arXiv.org · 0 citations