Multimodal Unlearning Evaluation Benchmark for VQA and CIFAR-10

Evaluation outputs for studying metric inconsistency in multimodal machine unlearning, supporting reproducibility for a NeurIPS 2026 paper. The dataset contains results on VQA benchmarks (MLLMU-Bench, UnLOK-VQA, MMUBench) and CIFAR-10 baseline results. It was created by author 'neurips26' and last updated on 2026-05-01.

Use Cases

• Reproducing results for the NeurIPS 2026 paper 'Metric Unreliability in Multimodal Machine Unlearning' based on the described evaluation outputs.
• Analyzing metric inconsistency in machine unlearning based on the provided benchmark results.
• Comparing unlearning performance across different VQA benchmarks (MLLMU-Bench, UnLOK-VQA, MMUBench) using the contained results.
• Establishing baseline comparisons using the CIFAR-10 results mentioned in the description.

Strengths

• Specifically supports reproducibility for a NeurIPS 2026 conference paper.
• Contains results across multiple established VQA benchmarks (MLLMU-Bench, UnLOK-VQA, MMUBench).
• Includes baseline results on the CIFAR-10 dataset as mentioned in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and sample data are unknown, which may limit suitability assessment.
• The license is unspecified, which could restrict usage.

Provenance

Source

huggingface

Collection Method

Likely generated as evaluation outputs for a research paper.

Time Range

The dataset itself appears to contain results from a specific research period, but the exact temporal coverage of the underlying data is not specified.

Freshness

Last updated 2026-05-01 08:45:11; freshness should be verified.

Geography

null