EML Munich

[ICML 2026] SOTAlign: Semi-Supervised Alignment of Unimodal Vision and Language Models via Optimal Transport

[CVPR 2026 Oral] FINER: MLLMs Hallucinate under Fine-grained Negative Queries

[ICML 2026] Sparse Autoencoders are Topic Models

[ICML 2026] The Latent Color Subspace: Emergent Order in High-Dimensional Chaos

[ICLR 2026] Are Reasoning LLMs Robust to Interventions on their Chain-of-Thought?

[NeurIPS 2025] Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models

[CVPR 2025] FLAIR: VLM with Fine-grained Language-informed Image Representations

[ECCV 2026] Training-free Uncertainty Guidance for Complex Visual Tasks with MLLMs