ECCV 2026 talks: diagnosing multimodal LLM contradictions, modality bias and OOD detection
abursuc · x · 2026-09-08
Live from ECCV 2026, the author highlights two conference talks:
- An opening talk by Gianni Franchi, "What Geometry Tells Us About OOD Detection?", examining out-of-distribution detection through a geometric lens.
- A talk by Teodora Gagaleska, "Diagnosing Multimodal LLMs: Contradictions, Modality Bias, and What Benchmarks Miss", digging into self-contradictions in multimodal LLMs, modality bias, and blind spots in current benchmarks.
Useful signal for anyone tracking multimodal model diagnostics and evaluation methodology.
Related event: ECCV 2026 workshop covers LLM uncertainty and multimodal diagnosis(3 posts)→
More from Research
- Mathematician argues papers should shift from proving conjectures to sharing insight — tak3sh8 · 2026-09-08
- Niantic's AutoCompass trains accurate visual localization from noisy GPS labels — ducha_aiki · 2026-09-08
- Apex unveils Astra, an automated AI research system aimed at AI improving AI — thetripathi58 · 2026-09-08
- More Ukrainian teams join CV research map with USDZ-prediction approach to object alignment and animation — ducha_aiki · 2026-09-08
- JustRL II adds token-level critic to GRPO for 128k long-CoT RL, lifting 2B model AIME25 from 61 to 81 — zibuyu9 · 2026-09-08
- Helsinki's Luigi Acerbi recruiting PhD/postdoc on amortized probabilistic ML for decision-making — LucaAmb · 2026-09-08