datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
memgym-dr-instances
MemGym-DR — Deep Research Benchmark Instances
Description
MemGym-DR is a multi-hop deep-research benchmark designed to stress-test long-horizon memory strategies in LLM agents. Each instance presents a multi-hop question whose sub-questions must be answered sequentially across turns; correct final answers require the agent to retain and synthesize information accumulated over many turns. Facts are drawn from 2WikiMultihopQA and augmented with synthetic distractors… See the full description on the dataset page: https://huggingface.co/datasets/MemGym/memgym-dr-instances.LISA_Plus_Instance_Seg
LISA++: An Improved Baseline for Reasoning Segmentation with Large Language Model
🤗Data | 📄Paper |
🚀Code | 💻Model |
🔥Citation
Dataset Details
Dataset type:
The LISA++ Instance Segmentation dataset is a QA dataset designed to train MLLM models for instance segmentation. It is based on the COCO2017 dataset.
Where to send questions or comments about the dataset:
https://github.com/dvlab-research/LISA
Paper:https://arxiv.org/abs/2312.17240
This model could be used… See the full description on the dataset page: https://huggingface.co/datasets/Senqiao/LISA_Plus_Instance_Seg.instance-level-tofu-unlearning
Instance-Level TOFU Benchmark
This dataset provides an instance-level adaptation of the TOFU (Maini et al, 2024) dataset for evaluating in-context unlearning in large language models (LLMs). Unlike the original TOFU benchmark, which focuses on entity-level unlearning, this version targets selective memory erasure at the instance level — i.e., forgetting specific facts about an entity.
It is compatible for evaluation with the locuslab/tofu_ft_llama2-7b model, which was fine-tuned on… See the full description on the dataset page: https://huggingface.co/datasets/chowfi/instance-level-tofu-unlearning.