Research map / Continual learning / Online and replay-based
Calibrated Knowledge Aggregation in Bayesian Mixture-of-Experts for Continual VQA
Mahsa Mozaffari, Hitesh Sapkota, Yu Kong, Xumin Liu, Qi Yu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Bayesian aggregation to task-ID router (ablation) | MoE router · VQA-v2 | accuracy: 57.2 -> 60.3 | Improved |
| add Bayesian aggregation to task-ID router (ablation) | MoE router · VQA-v2 | ECE: 0.15 -> 0.10 | Improved |
| add Bayesian aggregation to task-ID router (ablation) | MoE router · TDIUC-VS | ECE: 0.15 -> 0.07 | Improved |
| add utility-trained router instead of task-ID router (ablation) | MoE router · VQA-v2 | accuracy: 57.2 -> 63.25 | Improved |
| add utility-trained router instead of task-ID router (ablation) | MoE router · VQA-v2 | average forgetting: 3.0 -> 0.5 | Improved |
| add utility-trained router alone (ablation) | MoE router · VQA-v2 | router entropy: 0.84 -> 0.03 | Hurt |
| add utility-trained router alone (ablation) | MoE router · VQA-v2 | ECE: 0.15 -> 0.15 | No clear effect |
| add entropy regularization to utility router (ablation) | MoE router · VQA-v2 | router entropy: 0.03 -> 0.93 | Improved |
| add Bayesian aggregation to entropy-regularized utility router (ablation) | MoE router · VQA-v2 | ECE: 0.12 -> 0.07 | Improved |
| add Bayesian aggregation to entropy-regularized utility router (ablation) | MoE router · VQA-v2 | accuracy: 63.92 -> 64.16 | Improved |
| use utility router instead of oracle task-ID routing | MoE router · VQA-v2 | accuracy: 62.20 -> 63.92 | Improved |
| add replay memory (|B|>0) (ablation) | router · VQA-v2 | accuracy and ECE | Improved |
Closest papers by meaning
- SCoT: Similarity-guided Conflict-aware Task Consolidation for Continual VQA (ECCV 2026 · similarity 0.92)
- MacVQA: Adaptive Memory Allocation and Global Noise Filtering for Continual Visual Question Answering (AAAI 2026 · similarity 0.92)
- Re-evaluating Continual VQA: Toward Fair and Robust Evaluation for Multimodal Continual Learning (CVPR 2026 · similarity 0.91)
- Attention-Logit Steering to Compositional Generalization for Continual VQA (ECCV 2026 · similarity 0.90)
- CL-MoE: Enhancing Multimodal Large Language Model with Dual Momentum Mixture-of-Experts for Continual Visual Question Answering (CVPR 2025 · similarity 0.90)
- AVQACL: A Novel Benchmark for Audio-Visual Question Answering Continual Learning (CVPR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
