sensitivity
GSD-Sensitivity-Taxonomy-Labels
GSD-Sensitivity Taxonomy: Task Labels for Remote Sensing VQA
Per-task D / M1 / M2 taxonomy labels, inter-annotator agreement (IAA) data, and
evaluation traces for four public RS-VQA benchmarks.
Companion to *G. Park and D.-H. Lee, "Identifying the Measurement Gap in Remote
Sensing VQA with a GSD-Sensitive Taxonomy," IEEE Geosci. Remote Sens. Lett., 2026*
— accepted, DOI to follow. Code: github.com/ganghyunnnn/GSD-Sensitivity-Taxonomy
⚠️ This dataset contains annotations and… See the full description on the dataset page: https://huggingface.co/datasets/ganghyunnnn/GSD-Sensitivity-Taxonomy-Labels.usage-sensitivity-probe
usage-sensitivity-probe
Pair-distilled simulation of usage-dependent mention validity, mined from
rafmacalaba/data-use-mentions + Luna tier verdicts
(training/build_usage_sensitivity_sim.py, seed 0).
Every row contains a contrastive surface string — a mention judged BOTH as
a data source (tier1/tier2) in some contexts and as invalid
(tier3_nonmention/junk: promissory, logframe, container, bibliography, ...)
in others. Gold labels ONLY the data-source instances; activity… See the full description on the dataset page: https://huggingface.co/datasets/rafmacalaba/usage-sensitivity-probe.Arena-DROID-Camera-Sensitivity-Workflow-Sample
Arena DROID Camera Sensitivity Workflow Sample
Dataset Description
Arena-DROID-Camera-Sensitivity-Workflow-Sample is a compact set of episode-level results generated by an Isaac Lab-Arena simulation experiment. It lets users run the documented camera sensitivity analysis without first executing the policy-evaluation sweep.
The experiment evaluates an OpenPI pi05 policy on a DROID Rubik's-cube pick-and-place task while independently varying the wrist-camera… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Arena-DROID-Camera-Sensitivity-Workflow-Sample.alphastack-cost-sensitivityLlama-3_3-Nemotron-Super-49B-v1_5-sensitivity-maps
Llama-3_3-Nemotron-Super-49B-v1_5-sensitivity-maps
This dataset carries the per-layer quantization sensitivity maps of
nvidia/Llama-3_3-Nemotron-Super-49B-v1_5.
vramfit measured them.
A sensitivity map records one damage number per layer group and
candidate precision. Damage is the shift in the model's output
distribution when that group alone quantizes — mean final-logits KL
divergence against the bf16 reference. The maps describe the base
model, not any quantized file. They… See the full description on the dataset page: https://huggingface.co/datasets/Alberto-Codes/Llama-3_3-Nemotron-Super-49B-v1_5-sensitivity-maps.gemma-4-31B-it-sensitivity-maps
gemma-4-31B-it-sensitivity-maps
This dataset carries the per-layer quantization sensitivity map and
the in-frame importance matrix of
google/gemma-4-31B-it,
measured on the
QAT unquantized checkpoint
at revision 1e4d8beecacb8b7590c1d8bedd7335f687bf311f.
vramfit measured them.
A sensitivity map records one damage number per layer group and
candidate precision. Damage is the shift in the model's output
distribution when that group alone quantizes — mean final-logits KL
divergence… See the full description on the dataset page: https://huggingface.co/datasets/Alberto-Codes/gemma-4-31B-it-sensitivity-maps.